PHYSIOLOGICAL ADAPTATION TO LIMITATION IN A MARINE OLIGOTROPHIC ULTRAMICROBACTERIUM Sphingopyxis alaskensis

MARTIN OSTROWSKI

A thesis submitted in fulfilment of the requirements for the degree of Doctor of Philosophy

School of Biotechnology and Biomolecular Sciences Faculty of Science The University of New South Wales, Australia

October 2006 UNIVERSITY OF NEW SOUTH WALES Thesis/Project Report Sheet

Surname or Family name: OSTROWSKI First name: MARTIN Other name/s: - LUKE Abbreviation for degree as given in the University calendar: PhD School:BIOTECHNOLOGY AND BIOMOLECULAR SCIENCES Faculty: SCIENCE Title: PHYSIOLOGICAL ADAPTATION TO NUTRIENT LIMITATION IN A MARINE OLIGOTROPHIC ULTRAMICROBACTERIUM Sphingopyxis alaskensis

Abstract 350 words maximum:

Sphingopyxis (formerly ) alaskensis, a numerically abundant species isolated from Alaskan waters and the North Sea represents one of the only pure cultures of a typical oligotrophic ultramicrobacterium isolated from the marine environment. In this study, physiological and molecular characterization of an extinction dilution isolate from the North Pacific indicate that it is a strain of Sphingopyxis alaskenis, extending the known geographical distribution of this strain and affirming its importance as a model marine oligotroph. Given the importance of open systems in climatic processes, it is clearly important to understand the physiology and underlying molecular biology of abundant species, such as S. alaskensis, and to define their role in biogeochemical processes.

S. alaskensis is thought to proliferate by growing slowly on limited concentrations of substrates thereby avoiding outright starvation. In order to mimic environmental conditions chemostat culture was used to study the physiology of this model oligotroph in response to slow growth and nutrient limitation. It was found that the extent of nutrient limitation and starvation has fundamentally different consequences for the physiology of oligotrophic ultramicrobacteria compared with well-studied copiotrophic (Vibrio angustum S14 and Escherichia coli). For example, growth rate played a critical role in hydrogen peroxide resistance of S. alaskensis with slowly growing cells being 10, 000 times more resistant than fast growing cells. In contrast, the responses of V. angustum and E. coli to nutrient availability differed in that starved cells were more resistant than growing cells, regardless of growth rate.

In order to examine molecular basis of the response to general nutrient limitation, starvation and oxidative stress in S. alaskensis we used proteomics to define differences in protein profiles of chemostat-grown cultures at various levels of nutrient limitation. High-resolution two-dimensional electrophoresis (2DE) methods were developed and 2DE protein maps were used to define proteins regulated by the level of nutrient limitation. A number of these proteins were identified with the aid of mass spectrometry and cross-species database matching. The identified proteins are involved in fundamental cellular processes including protein synthesis, protein folding, energy generation and electron transport, providing an important step in discovering the molecular basis of oligotrophy in this model .

Declaration relating to disposition of project report/thesis

I am fully aware of the policy of the University relating to the retention and use of higher degree project reports and theses, namely that the University retains the copies submitted for examination and is free to allow them to be consulted or borrowed. Subject to the provisions of the Copyright Act 1968, the University may issue a project report or thesis in whole or in part, in photostat or microfilm or other copying medium.

I also authorise the publication by University Microfilms of a 350 word abstract in Dissertation Abstracts International (applicable to doctorates only).

...... Signature Witness Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing to the Registrar. Requests for a longer period of restriction may be considered in exceptional circumstances if accompanied by a letter of support from the Supervisor or Head of School. Such requests must be submitted with the thesis/project report.

FOR OFFICE USE ONLY Date of completion of requirements for Award

Registrar and Deputy Principal

Copyright and DAI Statement

‘I hereby grant the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all proprietary rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation.

I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral thesis only).

I have either used no substantial portions of copyright material in my thesis or I have obtained permission to use copyright material; where permission has not been granted I have applied/will apply for a partial restriction of the digital copy of my thesis or dissertation.'

Signature: ______Date: ______

Authenticity Statement

‘I certify that the Library deposit digital copy is a direct equivalent of the final officially approved version of my thesis. No emendation of content has occurred and if there are any minor variations in formatting, they are the result of the conversion to digital format.’

Signature: ______Date: ______

Table of Contents I

List of Figures V

List of Tables VII

List of Abbreviations VIII

Certificate of Originality X

Acknowledgements XI

Publications arising from this work XII

ABSTRACT XIV 1 GENERAL INTRODUCTION ...... 1 1.1 INTRODUCTION ...... 2 1.2 OLIGOTROPHIC ENVIRONMENTS ...... 3 1.2.1 General observations...... 3 1.2.1 An overview of the marine microbial ...... 4 1.2.2 Factors affecting microbial growth...... 8 1.3 DISTINCTIONS AND DEFINITIONS ...... 12 1.3.1 The size of small ...... 13 1.3.2 vs Oligotrophs...... 14 1.4 OLIGOTROPHIC ISOLATES ...... 16 1.4.1 Predicted properties of oligotrophs ...... 16 1.4.2 The extinction dilution method for isolating oligotrophs ...... 17 1.5 DESCRIPTION OF SPHINGOMONAS ALASKENSIS RB2256...... 21 1.5.1 Isolation of S. alaskensis...... 21 1.5.2 General growth characteristics ...... 23 1.5.3 Starvation and stress resistance...... 25 1.6 HYPOTHESIS ...... 26 1.6.1 Overall objectives...... 27 1.6.2 Specific aims ...... 27 2 MOLECULAR AND CHEMOTAXONOMIC CHARACTERISATION OF AN ABUNDANT ULTRAMICROBACTERIUM ISOLATED FROM THE OLIGOTROPHIC NORTH PACIFIC...... 29 2.1 BACKGROUND ...... 30 2.2 MATERIALS AND METHODS...... 32 2.2.1 Bacterial strains and culture conditions ...... 32 2.2.2 Alcohol precipitation...... 32 2.2.3 Estimation of concentration ...... 32 2.2.4 Restriction enzyme digestion...... 33 2.2.5 Agarose gel electrophoresis and visualisation ...... 33 2.2.6 Genomic DNA extraction...... 34 2.2.7 Amplification of 16S rDNA sequences ...... 34 2.2.8 Purification and nucleotide sequencing of amplification products ...... 35 2.2.9 Phylogenetic analysis of 16S rDNA sequences...... 35 2.2.10 DNA-DNA hybridisation and mol%G+C content analysis ...... 35 2.2.11 Fatty acid analysis...... 36 2.3 RESULTS...... 38 2.3.1 Sequencing and phylogeny of strain AFO1 ...... 38 2.3.2 DNA base composition and DNA-DNA hybridisation ...... 42 2.3.3 Fatty acids...... 43 2.4 DISCUSSION...... 45 2.4.1 Ecological implications of S. alaskensis sp. strain AFO1...... 45 2.4.2 Reclassification of Sphingomonas alaskensis to Sphingopyxis alaskensis 47 2.4.3 Conclusion ...... 48 3 CONTINUOUS CULTURE METHODS FOR INVESTIGATING PHYSIOLOGICAL RESPONSES TO NUTRIENT LIMITATION IN MARINE BACTERIA...... 50 3.1 BACKGROUND ...... 51 3.1.1 Small scale chemostats ...... 56 3.1.2 Theoretical considerations: growth rates and nutrient flux...... 57 3.2 MATERIALS AND METHODS...... 59 3.2.1 Bacterial strains, media and culturing conditions...... 59 3.2.2 Total counts and morphology...... 60 3.2.3 Viability measurements...... 61 3.3 RESULTS...... 62 3.3.1 Validation of mini chemostats...... 62 3.3.2 Substrate limitation ...... 63 3.3.3 Steady-state growth and viability of S. alaskensis and V. angustum S14 in chemostat culture...... 64 3.3.4 Morphology and cell dimensions ...... 67 3.4 DISCUSSION...... 71 3.4.1 The impact of nutrient limitation on the viability, morphology and cell dimensions of S. alaskensis and V. angustum S14...... 71 3.4.2 Growth yield and maintenance energy of S. alaskensis and V. angustum S14 in response to nutrient limitation...... 74 4 PHYSIOLOGICAL RESPONSES TO NUTRIENT LIMITATION AND STRESS FOR AN OLIGOTROPHIC ULTRAMICROBACTERIUM, SHINGOPYXIS ALASKENSIS...... 77

4.1 BACKGROUND ...... 78 4.2 MATERIALS AND METHODS...... 81 4.2.1 Bacterial strains, maintenance and culturing conditions...... 81 4.2.2 Viability measurements...... 81 4.2.3 Stress exposure experiments ...... 82 4.2.4 Determination of catalase activity ...... 82 4.3 RESULTS...... 84 4.3.1 The effect of growth rate on hydrogen peroxide resistance in S. alaskensis...... 84 4.3.2 The effect of temperature on peroxide resistance ...... 86 4.3.3 The effect of carbon vs nitrogen limitation on peroxide resistance ...... 88 4.3.4 The effect of growth rate on hydrogen peroxide resistance in V. angustum S14...... 90 4.3.5 The effect of growth rate on hydrogen peroxide resistance in E. coli....92 4.3.6 Catalase activity and hydrogen peroxide resistance in S. alaskensis....94 4.3.7 The effect of growth rate on resistance to ultraviolet radiation in S. alaskensis ...... 97 4.3.8 The effect of growth rate on heat stress and freeze-thaw resistance in S. alaskensis ...... 98 4.3.9 The possible role of cellular pigments in stress resistance of S. alaskensis ...... 103 4.3.10 The effect of growth conditions and nutrient limitation on the of nostoxanthin of S. alaskensis ...... 104 4.4 DISCUSSION...... 107 4.4.1 Growth rate control of hydrogen peroxide resistance ...... 107 4.4.2 Possible mechanisms of growth rate control of hydrogen peroxide resistance...... 109 4.4.3 Resistance to ultraviolet radiation ...... 112 4.4.4 Starvation vs low growth rate induction of peroxide stress resistance 113 4.4.5 Possible role of cell membrane and pigments in overall stress resistance 114 5 AN ASSESSMENT OF TWO DIMENSIONAL PROTEIN REFERENCE MAPS FROM S. ALASKENSIS RB2256 ...... 118 5.1 BACKGROUND ...... 119 5.2 MATERIALS AND METHODS...... 122 5.2.1 sampling and preservation...... 122 5.2.2 Pulse labelling of chemostat cultures...... 122 5.2.3 Sample preparation ...... 122 5.2.4 Isoelectric focusing...... 123 5.2.5 Equilibration and SDS-PAGE...... 123 5.2.6 Gel staining...... 124 5.2.7 Image acquisition and spot detection...... 124 5.3 RESULTS AND DISCUSSION ...... 127 5.3.1 Characteristics of silver stained and radiolabelled 2DE images ...... 127 5.3.2 Comparison of silver stained and radiolabelled 2DE images...... 130 5.3.3 Possible explanations for quantitative differences between methods ..131 5.3.4 Representation of the S. alaskensis proteome in the pH 4-7, Mr 10-200 kDa window...... 135 5.3.5 Conclusion ...... 139 6 EXAMINATION OF GENE EXPRESSION IN RESPONSE TO NUTRIENT LIMITATION IN S. ALASKENSIS...... 140 6.1 BACKGROUND ...... 141 6.2 MATERIALS AND METHODS ...... 146 6.2.1 Two-dimensional electrophoresis, image acquisition and analysis.....146 6.2.2 Preparative 2DE and semi-dry blotting ...... 146 6.2.3 Blot staining...... 146 6.2.4 Sample preparation and in-gel trypsin digestion...... 147 6.2.5 Mass spectrometry analysis ...... 148 6.2.6 Amino acid and Edman degradation...... 149 6.3 RESULTS...... 150 6.3.1 Differential expression of proteins observed on silver stained gels ....150 6.3.2 Differential expression of pulse labelled proteins ...... 155 6.3.3 Identification of protein spots on the basis of peptide mass fingerprinting and Edman sequencing...... 160 6.3.4 Identification of protein spots by LC-ESI MS/MS...... 162 6.4 DISCUSSION...... 165 6.4.1 Differential expression of proteins in steady state chemostats...... 165 6.4.2 Cross species identification ...... 170 6.4.3 Identified proteins ...... 172 7 GENERAL DISCUSSION...... 176 7.1 SUMMARY ...... 177 7.2 EXTINCTION DILUTION, ...... 178 7.2.1 S. Alaskensis sp. strain AFO1...... 178 7.2.2 Extinction dilution and lab domestication...... 179 7.3 THE IMPACT OF NUTRIENT LIMITATION ON THE PHYSIOLOGY OF MARINE BACTERIA...... 182 7.3.1 Distinct strategies?...... 182 7.3.2 Comparison of S. alaskensis with Peligibacter ubique ...... 184 7.4 THE MECHANISMS OF STRESS RESISTANCE IN S. ALASKENSIS ...... 186 7.4.1 Carotenoids...... 188 7.5 GENE EXPRESSION DURING NUTRIENT LIMITED GROWTH AND STARVATION ...... 189 7.6 FUTURE PROSPECTS...... 193 7.6.1 Molecular genetics ...... 193 7.6.2 Cross species identification of proteins: environmental proteomics ...194 7.7 CONCLUDING REMARKS ...... 195 8 REFERENCES ...... 197 LIST OF FIGURES

Figure 2.1 Distance-matrix tree of selected Sphingomonas 16S rDNA 41 sequences

Figure 3.1 Schematic diagram of the chemostat apparatus used in this study 55

Figure 3.2 Development of total cell numbers and CFU in a glucose limited 66 chemostat of S. alaskensis

Figure 3.3 Molar growth yield and cell numbers of S. alaskensis and V. 67 angustum S14 grown in glucose limited chemostats at different specific rates of growth

Figure 4.1 Percent survival of S. alaskensis grown at different specific growth 85 rates in glucose-limited chemostats after exposure to 25mM hydrogen peroxide for 60 min

Figure 4.2 Percent survival of nutrient limited chemostat grown S. alaskensis 87 cells following exposure to 25 mM hydrogen peroxide for up to 60 min

Figure 4.3 Percent survival of S. alaskensis in response to the concentration of 89 hydrogen peroxide

Figure 4.4 Percent survival of nutrient limited chemostat grown S. alaskensis 91 cells following exposure to 25 mM hydrogen peroxide for up to 60 min

Figure 4.5 Percent survival of V. angustum S14 following exposure to 2mM 93 hydrogen peroxide for up to 60 min after growth at various rates in glucose limited chemostats

Figure 4.6 Percent survival of log-phase, starved and chemostat grown E. coli 95 K12 UNSW002900 after exposure to 25 mM hydrogen peroxide for up to 60 min

Figure 4.7 Survival of batch grown S. alaskensis cells following exposure to 99 UV radiation. UV doses ranged from 0 to 10,000 J m-2

Figure 4.8 Percent survival of batch and chemostat grown S. alaskensis 101 following exposure to 4000 Jm-2 ultraviolet radiation

Figure 4.9 Percent survival of exponentially growing S. alaskensis following 102 heat stress

Figure 4.10 Absorbance spectra of cellular methanol extracts of S. alaskensis 105

Figure 4.11 Molecular structure of nostoxanthin 105

Figure 4.12 Relative abundance of carotenoids extracted from S. alaskensis 106 grown in glucose-limited chemostats and batch cultures

35 Figure 5.1 The incorporation of [ S] L-methionine for cultures grown at 129 35 Figure 5.1 The incorporation of [ S] L-methionine for cultures grown at 129 different glucose limited rates in chemostat cultures

Figure 5.2 Comparison of protein gel images prepared from 2DE profiles of 134 silver-stained total protein and radioactively labelled proteins of S. alaskesnis grown in chemostat culture

Figure 5.3 Distinct features present in radiolabelled and silver stained gel 135 images

Figure 5.4 Theoretical 2DE protein maps for - closely related 138 to S. alaskensis

Figure 6.1 Outline of protein identification strategy on the basis of micro- 145 Liquid Chromatography Electrospray Ionisation Tandem Mass Spectrometry (LC-ESI MS/MS)

Figure 6.2 A comparison of silver-stained 2DE gel images of total protein 152 taken from S. alaskensis grown in glucose limited chemostats

Figure 6.3 Differential intensity for silver stained protein spots from S. 154 alaskensis cultures grown at low relative to high rates of growth in glucose-limited chemostats

Figure 6.4 A comparison of 2DE gel images of protein from S. alaskensis 157 grown in glucose limited chemostats

Figure 6.5 Differential intensity for radiolabeled protein spots from S. 159 alaskensis cultures grown at low relative to high rates of growth in glucose-limited chemostats

Figure 6.6 Two dimensional reference image showing the protein spots 169 characterised and their regulation grouping LIST OF TABLES

Table 1.1 Description of terms relating to the small size of bacteria 16

Table 1.2 General characteristics of selected marine oligotrophic isolates 20

Table 2.1. Changes and gaps relative to S. alaskensis RB2256T 16S rDNA 40 sequence

Table 2.2 DNA base composition and DNA:DNA hybridisation values 42 between S. alaskensis RB2256, strains AFO1and KT-1, and S. macrogoltabidus

Table 2.3 FAME analysis of fatty acids from S. alaskensis RB2256 and 44 Sphingomonas strains AFO1 and KT-1

Table 3.1 Assumed physiological constants and their relationships during 56 steady state cultivation in chemostat culture

Table 3.2 Mixing time, a measure of homogeneity in ‘classic’ large scale and 63 small scale chemostat culture vessels

Table 3.3 Size and volume measurements for S. alaskensis determined by 70 transmission electron Microscopy and phase contrast microscopy

Table 3.4 Size and volume measurements for V. angustum S14 determined 70 from estimates based on phase contrast microscopy

Table 4.1 Catalase activity measured in whole cells of S. alaskensis grown in 98 glucose-limited chemostats and batch cultures

Table 5.1 A comparison of the features of 2DE reference maps generated by 130 silver staining of total protein and pulse labelling of newly synthesised proteins with radioactive methionine

Table 6.1 Examples of the number of unique, differentially expressed and total 143 proteins analysed in selected bacterial 2DE projects

Table 6.2 Summary of the number of features detected and the number of spots 151 analysed from replicate silver-stained and radiolabelled gel images

Table 6.3 Summary of characteristics for differentially expressed protein spots 153 on silver stained 2DE gels

Table 6.4 Summary of characteristics for differentially expressed protein spots 158 on silver radiolabeled 2DE gels

Table 6.5 Summary of differentially expressed protein spots on silver stained 160 and radiolabelled 2DE gels

Table 6.6 Proteins characterised by mass spectrometry or Edman sequencing 164 from the S. alaskensis proteome LIST OF ABBREVIATIONS

2DE Two Dimensional gel Electrophoresis A Adenosine ADP adenosine diphosphate APAF AustralianProteomeAnalysis Facility ASW Artificial Sea Water ASW-G Artificial Sea Water-Glucose ATP adenosine triphosphate ATP-ase Adenosine triphosphate synthase BCA Bicinchoninic Acid BCIP 5-bromo-4-chloro-3-indolyl BLAST Basic Local Alignment Search Tool bp base pair(s) cAMP cyclic Adenosine Mono Phosphate CBB Coomasie Brilliant Blue CFU Colony Forming Units CHAPS 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate Ci Curie CID Collision-Induced Dissociation CRP cAMP Receptor Protein CS-screen Carbon Sulphur-screen CTAB Hexadecyltrimethylammoniumbromide d day(s) Dia. diameter db database DEPC diethyl-pyrocarbonate DFAA dissolved free amino acids DI Differential Intensity DNA Deoxyribonucleic acid dNTP deoxy nucleotide triphosphate DOC Dissolved Organic Carbon dpm disintegration per minute DTT Dithiothreitol EDTA Ethylene diamine tetraacecitc acid, trisodiumsalt ESI-MS Electron Spray Ionization-Mass Spectrometry EtBr Ethidium Bromide FAS Filtered-Autoclaved Seawater FISH Fluorescent In Situ Hybridation g gram(s) GSL glycosphingolipid GTP Guonisine triphosphate h hour(s) HPLC High Performance Liquid Chromatography IEF IsoElectric Focusing IPG Immobilized pH Gradient J Joules kbp kilo basepair(s) kDa kiloDalton(s) kV kilo Volt LB Luria Bertani μLC-ESI MS Liquid Capillary-ElectronSpray IonisationMass Spectrophotometer LMP Low Melting Point MALDI-TOF Matrix Assisted Laser Disorption Ionisation – Time of Flight Mb mega basepairs Min Minutes MNB Marine Nutrient Broth MOWSE Molecular Weight Search MS Mass Spectrometry Mr molecular we ig h t NCBI National Center for Biotechnology Information nm nanometer(s) nrdb non redundant database OD Optical Density

OD433 Optical Density at 433 nm

OD610 Optical Density at 610 nm PAGE PolyAcrylamide Gel Electrophoresis PBS Phosphate buffer saline PCR Polymerase Chain Reaction pI Isoelectric point PMF Peptide Mass Fingerprinting Ppm partspermillion PVDF polyvinylidine difluoride RNA Ribonucleic acid rpm revolutions per minute SDS sodium dodecyl sulphate s second(s) sp. species TAE Tris-Acetic acid-EDTA TCA Trichloroacetic acid TE Tris-EDTA TEMED N,N,N’,N’-Tetramethylethylene-diamine TOF Time of Flight UV-B Ultra Violet –B (wavelengths, 290-320 nm) Vh volt-hours CERTIFICATE OF ORIGINALITY

I hereby declare that this submission is my own work and to the best of my knowledge it contains no material previously published or written by another person, nor material which to a substantial extent has been accepted for the award of any other degree or diploma at UNSW or any other education institution, except where due acknowledgement is made in the thesis. Any contribution made to the research by others, with whom I have worked in UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual context of this thesis is the product of my own work, except to the extent that assistance from others in the project’s design and conception and linguistic expression is acknowledged.

Martin Ostrowski ACKNOWLEDGEMENTS

This thesis would not have been possible without the contributions of many people, for which I am deeply indebted. I would like to sincerely thank my supervisor Associate Professor Rick Cavicchioli for his guidance, undersdtanding, patience and constant encouragement. In addition I would also like to thank all of my supervisors, past and present, Jan Gottschal, Tom Stewart, Dave Scanlan and Paul March for fostering my love of science and sharing my successes and failures. Thank you to Dr. Valerie Wasinger and Dr. Garry Corthals, formerly of the Protein Lab at the Garvan Institute, for their assistance and kind advice in performing ESI-MS analysis. To Dr John Bowman for DNA-DNA hybridisation and Dr David Nichols for fatty acid analysis. I must especially acknowledge my colleagues Fitri Fegatella and Amber Goodchild for their insight, cooperative assistance and thoughtful discussions. Thanks to Dr Neil Saunders for teaching me how to use a computer. Thanks to other members of the laboratory, Julie Lim, Torsten Thomas, Nuria, Laura and Sohail for their support and encouragment. Thanks also to Caiyan, Anne-Carlijn, Andrew, Phil, Michael, Kristine and Christine for being such lovely and enthusiastic students. Parts of this work were supported by the Australian Research Council and the Department of Science, Industry and Tourism. My subsistence was supported by an Australian Postgraduate Award, the Australian Society for Microbiology Research Trust Award, the School of Microbiology and Immunology and my supervisor, and this is duly acknowledged. I am truly grateful for all of the great relationships that I’ve had with people throughout the faculty of Life Science and beyond. Thanks to Shaun for too many things to mention here. Special mention to my great friend Charlie, I will truly miss the many times we’ve been out drinking and the many times we’ve shared our thoughts. I wonder if they ever truly noticed the kind of thoughts we’ve got? Thanks to my parents and extended family who have always projected their comforting familial warmth and support from various corners of the World. Lastly, to Sophie for loyalty, loving support and always being there in times of need. PUBLICATIONS ARISING FROM THIS WORK

1. Ostrowski, M., F. Fegatella, V. Wasinger, M. Guilhaus G.L. Corthals, and R. Cavicchioli. 2004. Cross-species identification of proteins from proteome profiles of the marine oligotrophic ultramicrobacterium, Sphingopyxis alaskensis. Proteomics 4:1779-1788.

2. Cavicchioli. R. and M. Ostrowski. 2003. Ultramicrobacteria. Encyclopedia of Life Sciences. Publishing Group, London

3. Cavicchioli R., M. Ostrowski, F. Fegatella, A. Goodchild, N. Guixa- Boixereu. 2003. Life Under Nutrient Limitation in Oligotrophic Marine Environments: an Eco/Physiological Perspective of Sphingopyxis alaskensis (formerly Sphingomonas alaskensis). Microbial 45:203-217.

4. Ostrowski, M., R. Cavicchioli, M. Blaauw, and J.C. Gottschal. 2001. Specific growth rate plays a critical role in hydrogen peroxide resistance of the marine oligotrophic ultramicrobacterium Sphingomonas alaskensis strain RB2256. Applied and Environmental Microbiology 67: 1292-1299.

5. Eguchi, M., M. Ostrowski,F.Fegatella,J.Bowman,D.Nichols,T.Nishino and R. Cavicchioli. 2001. Sphingomonas alaskensis strain AF01: an abundant oligotrophic ultramicrobacterium form the North Pacific. Applied and Environmental Microbiology 67:4945-4954.

6. Fegatella, F., M. Ostrowski and R. Cavicchioli. 1999. An assessment of protein profiles from the marine oligotrophic ultramicrobacterium, Sphingomonas sp., strain RB2256. Electrophoresis 20: 2094-2098.

7. Cavicchioli, R., F. Fegatella, M. Ostrowski, M. Eguchi and J. Gottschal. 1999. Sphingomonads from marine environments. Journal of Inductrial Microbiology and Biotechnology 23: 268-272. CONFERENCE PROCEEDINGS

1. Eguchi, M., M. Ostrowski, F. Fegatella, J. Bowman, D. Nichols, T. Nishino, and R. Cavicchioli. 2001. Sphingomonas alaskensis strain AF01: an abundant oligotrophic ultramicrobacterium form the North Pacific. Poster. 9th International Society of Congress, Amsterdam, Netherlands. August 25-31.

2. Cavicchioli, R., M. Ostrowski, and F. Fegatella. 1999. Microbial physiology of a model oligotrophic marine ultramicrobacterium. 9th International Congress of Bacteriology and Applied Microbiology at the International Union of Microbiological Societies (IUMS), Sydney, August 16-20.

3. Ostrowski, M., R. Cavicchioli, M. Blaauw, and J.C. Gottschal. 2001. Growth-rate control of hydrogen peroxide resistance of the marine oligotrophic ultramicrobacterium Sphingomonas alaskensis strain RB2256. 9th International Congress of Bacteriology and Applied Microbiology at the International Union of Microbiological Societies (IUMS), Sydney, August 16-20. ABSTRACT

Sphingopyxis (formerly Sphingomonas) alaskensis, a numerically abundant species isolated from Alaskan waters and the North Sea, represents one of very few pure cultures representative of oligotrophic ultramicrobacteria isolated from the marine environment. In this study, physiological and molecular characterization of an extinction dilution isolate from the North Pacific indicate that it is a strain of Sphingopyxis alaskenis, extending the known geographical distribution of this strain and affirming its importance as a model marine oligotroph. Given the importance of open ocean systems in climatic processes, it is clearly important to understand the physiology and underlying molecular biology of abundant species, such as S. alaskensis, and to define their role in biogeochemical processes.

S. alaskensis is thought to proliferate by growing slowly on limited concentrations of substrates thereby avoiding outright starvation. In order to mimic environmental conditions chemostat culture was used to study the physiology of this model oligotroph in response to slow growth and nutrient limitation. It was found that the extent of nutrient limitation and starvation has fundamentally different consequences for the physiology of oligotrophic ultramicrobacteria compared with well-studied copiotrophic bacteria (Vibrio angustum S14 and Escherichia coli). For example, growth rate played a critical role in hydrogen peroxide resistance of S. alaskensis with slowly growing cells being 10, 000 times more resistant than fast growing cells. In contrast, the responses of V. angustum and E. coli to nutrient availability differed in that starved cells being more resistant than growing cells, regardless of growth rate.

In order to examine the molecular basis of the response to general nutrient limitation, starvation and oxidative stress in S. alaskensis we used proteomics to define differences in protein profiles of chemostat-grown cultures at various levels of nutrient limitation. High-resolution two-dimensional electrophoresis (2DE) methods were developed and 2DE protein maps were used to define proteins regulated by the level of nutrient limitation. A number of these proteins were identified with the aid of mass spectrometry and cross-species database matching. The identified proteins are involved in fundamental cellular processes including protein synthesis, protein folding, energy generation and electron transport, providing an important step in discovering the molecular basis of oligotrophy in this model organism. Chapter 1

1GENERAL INTRODUCTION

1 Chapter 1

1.1 INTRODUCTION

Oceans have the highest cellular production rate of any on the planet and yet are vast nutrient-limited environments. The high level of is largely due to the phototrophic prokaryotic primary-producers, and the heterotrophic which effect nutrient transformation and remineralisation. The fixation of carbon, nitrogen and by marine bacteria, and their subsequent conversion into particulate matter are critically important processes in marine environments that form the basis for the grazing and sinking food chains of the . Heterotrophic ultramicrobacteria are major contributors to oceanic and terrestrial biogeochemical cycles (Whitman et al., 1998). As of in oligotrophic marine , they interact with all trophic levels and control the nutrient fluxes via mineralisation thus impacting on the productivity of all marine life. With predictions of increasing ocean oligotrophy as a consequence of global warming (Matear et al., 1999, Wignal et al., 1996) it is clearly important to understand the physiology of this class of bacteria in order to determine the impact they have on oceanic .

A major portion of terrestrial and marine environments are oligotrophic (nutrient poor) and thus, the day-to-day situation for almost all bacteria is that they are limited in their supply of one or more essential nutrients. Despite low levels of nutrients the pelagic marine environment is dominated, in terms of biomass and activity, by small bacteria, variously referred to as ultramicrobacteria, microcells, nanoplankton or depending on the terminology adopted by microbial ecologists and physiologists at different points in time (discussed below). This phylogenetically

2 Chapter 1

diverse bacterial size class is believed to proliferate by growing slowly and efficiently competing for limited concentrations of growth nutrients, thereby avoiding outright starvation. The small size, slowly growing bacterial cells that dominate oligotrophic environments are referred to as oligotrophic bacteria, or oligotrophs. Despite the relatively good understanding of the physiology and genetics of readily cultured, faster growing marine bacteria, termed copiotrophs, oligotrophic ultramicrobacteria and the specific roles they play in environmental processes are poorly understood. Knowledge about the physiology of oligotrophs is limited by the availability of environmental isolates, relating largely to the difficulty in culturing these bacteria from their environments and technical limitations for investigating characteristics of marine bacteria under ecologically relevant conditions. To date, most insight into the physiology of the truly low nutrient adapted marine bacteria has been obtained from very few strains, including

Sphingomonas alaskensis isolated by extinction dilution culture as a numerically dominant bacterium from Resurrection Bay, Alaska, and the North Sea (Schut et al.,

1993).

1.2 OLIGOTROPHIC ENVIRONMENTS

1.2.1 General observations

Oligotrophic environments generally lack exogenous supply of nutrients and are defined by a low nutrient flux (< 1 mg carbon per litre per day, Schut et al., 1997) as well as by low absolute concentrations of nutrients (Morita, 1997). According to this definition, a large proportion of the world’s oceans may be considered

3 Chapter 1

oligotrophic (Cole et al., 1988). Despite the low level of nutrients microbial numbers persist on the order of 105 -106 cells ml-1 in the upper 200m of the ocean

(the photic zone) where the bulk of activity occurs. Total numbers in the ocean are estimated at 1029 (Whitman et al., 1998), as a result, marine contribute a large proportion of the worlds biosphere in terms of carbon, nitrogen and phosphorus. Furthermore, of the three largest microbial (seawater, soil and sediment/soil subsurface), the rates of cellular activity and turnover are highest in the open ocean (Whitman et al., 1998). By virtue of their abundance and biomass heterotrophic prokaryotes in the ocean play an essential role in nutrient transformation and remineralisation. In addition, picophytoplankton (phototrophic prokaryotes and ), contribute significantly to global primary production (Li et al., 1992, Campbell et al., 1994, Li,

1994, Vaulot et al., 1995), with estimates as high as 50% of global carbon fixation attributed to this size class (Partensky et al., 1999). Thus, together the smallest heterotrophic and phototrophic cells play an essential role in regulating the accumulation, export, re-mineralisation and transformation of the world’s largest pool of organic carbon (Cole et al., 1988, Carlson et al., 1996) resulting in an ecosystem composed primarily of a microbial where prokaryotes and picoeukaryotes represent the most important biological component.

1.2.1 An overview of the marine microbial community

When observed directly, indigenous bacterial communities are rich in carbon and nitrogen and exhibit a low protein and DNA content, they display typical cell volumes in the range from 0.02 – 0.12 m3, around an order of magnitude smaller

4 Chapter 1

than commonly studied bacteria such as Escherichia coli (Fuhrman, 1981, Simon and Azam, 1989, Schut et al., 1993, Strehl et al., 1999, Button, 2000). Initial attempts to isolate marine bacteria on nutrient rich agar plates revealed a discrepancy of up to three orders of magnitude between plate counts and the observed total number of cells in marine samples (Jannasch and Jones, 1959, Ferguson et al., 1984). Taken together with the observation that a high proportion of early ocean isolates were typically larger (0.34-6.4 m3) and undergo starvation induced miniaturisation processes (Mården et al., 1985, Nyström et al., 1986, Lee and Fuhrman, 1987, Moyer and Morita, 1989a, Schut et al., 1993), it was believed that indigenous microcells represent starved and dormant forms of isolates that could not form colonies on agar plates. It was therefore assumed that in the environment starvation was the natural state of microorganisms. This assertion, however, does not account for a number of observed phenomena listed here. (i) On a per unit volume basis, oceanic microcells exhibit higher activity than the atypically large cells (Douglas et al., 1987, Eguchi and Ishida, 1990, Ouverney and Fuhrman, 1999). (ii) More than 90% of the productivity in pelagic regions is due to free-living, rather than substrate-attached, cells (Cho and Azam, 1988). (iii) Bacteria that remain small when actively growing have been observed and isolated (Ishida and Kadota, 1981, Schut et al., 1993, Rappé et al., 2002). (iv) The global significance and activity of ultramicrobacterial phototrophic is well established (Partensky et al., 1999), and (v)

Starved, or dormant, bacteria may not become predominant in the ocean while in the non-growing state.

More recently molecular techniques suggest low in situ abundance of typically isolated bacteria (Eilers et al., 2000). Developments with molecular methods

5 Chapter 1

enabled the relative abundance of specific prokaryotic taxa to be determined without the need to cultivate microorganisms. Initial studies based on SSU rRNA sequence libraries found that the most abundant rDNA sequences obtained did not correspond to cultured species and were distantly related to other rDNA sequences in databases

(reviewed in Giovannoni and Rappé, 2000). These results clearly demonstrated that natural bacterial communities were composed of unknown species that were incapable of forming colonies on commonly used microbiological media. As a consequence, laboratory studies with readily cultured marine bacteria cannot be regarded as a good representation of the numerically dominant indigenous species.

The profound ‘unculturability’ of the microbial community as a whole has severely limited our understanding of the natural state and role of microorganisms from this environment. As a consequence our current knowledge has been largely restricted to direct observation of microbial populations, measurements of in situ activity and the isolation and analysis of community DNA. It is important to note that without detailed characterisation of representative isolates the functional significance of abundant bacterioplankton remains unknown.

Together with rDNA sequencing data from clone libraries the abundance of specific bacterial groups has been estimated using epifluorescence microscopy aided by the development of species-specific probes and fluorescence in situ hybridisation

(Porter and Feig, 1980, and e.g. Glöckner et al., 1999, Cottrell and Kirchman, 2000,

West et al., 2001). A major outcome of molecular based studies is a realisation that the majority of bacterioplantkon belong to just 11 major phylogenetic groups

(Giovannoni and Rappé, 2000). More than 80% of all bacterial rDNA sequences obtained from marine samples belong to only nine phylogenetic groups, most of

6 Chapter 1

which have no known cultured representatives. Furthermore, uncultured members of the SAR11, SAR116 and Roseobacter lineages of the -Proteobacteria, account for the majority of all rRNA that have been identified in seawater while sequences account for about one quarter of the total (Rappé et al.,

2002). Further studies have established distinctions between the composition of coastal and open ocean waters (Fuhrman and Ouverney, 1998, Rappé et al., 2000,

Fuller et al., 2005), particulate-associated and free-living communities (e.g. DeLong et al., 1993) as well as the specific depth distributions of phylogenetic clades

(Moore et al., 1995, West and Scanlan, 2001).

More recently, methods have been used to estimate the relative contribution of actively growing bacteria in the total bacterial pool (Karner and Fuhrman, 1997,

Gasol et al., 1999). An approach recently developed combines microautoradiography and FISH (Lee et al., 1999, Ouverney and Fuhrman, 1999) and allows the uptake of dissolved organic matter (DOM) to be determined in phylogenetically distinct groups. This has been applied to the uptake of amino acids in marine (Ouverney and Fuhrman, 2000), and supports the observation that different bacterial taxa are responsible for the uptake of low- or high-molecular weight DOM (Cottrell and Kirchman, 2000).

It is important to recognize the limitations of current methods for assessing microbial diversity and activity. Representation in clone libraries is dependent upon methods for extracting community DNA while PCR and probe methods also involve certain biases associated with primer and probe sequences. Enumerating heterotrophic bacteria by microscopy is complicated by the small size of cells 7 Chapter 1

(reviewed in MacGregor, 1999). This causes problems with detection due to the limits of optical resolution, and discriminating heterotrophic cells from auto- fluorescing particles. This also becomes an issue when discriminating small, phototrophic bacteria such as spp. which are abundant in oligotrophic waters (Partensky et al., 1999), and can be difficult to distinguish from (Sieraki et al., 1995). Low cellular rRNA content (Fegatella et al.,

1998, Oda et al., 2000) and membrane impermeability also provide limitations for methods relying on FISH or dyes for detection (Pinhassi et al., 1997, Glöckner et al., 1999, Oda et al., 2000). The specific limitations inherent in current methods make it clear that multiple independent approaches are required for understanding the composition and dynamics of the .

1.2.2 Factors affecting microbial growth

The fate of prokaryotes in marine food webs has been linked to the grazing activities of nanoplankton organisms (3-20 μm, Sherr and Sherr, 1994, Caron et al., 1999) and to viral lysis (Fuhrman and Suttle, 1993, Furhman, 1999). Consequently, knowledge of the factors that control the abundance, biomass and activity of microorganisms in the open ocean is an essential consideration for the understanding of the biogeochemical fluxes in the ocean (Azam et al. 1984). Despite the fact that the ocean is continuous, it is composed of distinct macro-zones that are affected or defined by currents, stratification, mixing, water depth, photic, non-photic, light intensity, etc. Even within distinct zones, micro-zones persist as a result of microbial distribution, microbial interactions and the colloidal composition of seawater. As a result, microbial populations have evolved a dynamic that reflects the various levels

8 Chapter 1

of heterogeneity within what otherwise appear as a vast homogenous expanse.

Spatial variability includes depth, stratification, aggregates (marine snow) (Alldredge and Silver, 1988), bacteria attached to algae, and even smaller microstructures

(Azam, 1998). Strong inverse gradients of light and nutrients are observed in stratified water columns with intense for organic and inorganic nutrients at the surface and higher concentrations of nutrients in light limited areas at the bottom of and below the photic zone (Zubkov et al., 2001, Cavender-Bares et al.,

2001). Temporal variability is also important in some oligotrophic oceans. During the break down of thermoclines in temperate areas (Ducklow et al., 1993, Estrada,

1996, Ducklow, 1999), phytoplankton exhibit higher activities than during either the normal winter or summer periods due to the upwelling of nutrient rich waters to the surface. In tropical latitudes, seasonal variability may also be affected by climatic phenomena such as El Niño (Equatorial Pacific, Ducklow et al., 1995) and Monsoon

(Indian Ocean, Wiebinga et al., 1997). Support for this is illustrated by the distribution of Prochlorococcus spp. in the water column where light intensities vary by up to 4 orders of magnitude (Moore et al., 1995, Partensky et al., 1999, West and

Scanlan, 2001). Throughout this range, physiologically and genetically distinct populations of Prochlorococcus exist which have adapted to high- or low-light intensities (Moore et al., 1998). It has been argued that the co-existence and distribution of distinct eco-types allows the perpetuation of this genus over a wider range of growth conditions than would other wise occur for a single (Moore et al., 1998). Moreover another illustration of this point is that due to different nutrient acquisition abilities and cell requirements, the growth of co-existing microbial populations can be simultaneously limited by different nutrients, for

9 Chapter 1

example, limited by light and inorganic nutrients and heterotrophs limited by organic carbon (Thingstad and Lignell, 1997).

Not only are large numbers of sub-micron colloids present (Nagata and Kirchman,

1997) but recent experiments demonstrate the spontaneous polymerisation of marine

DOM into polymer gels (Chin et al., 1998). The emerging picture is “seawater as an organic matter continuum, a gel of tangled polymers with embedded strings, sheets, and bundles of fibrils and particles, including living organisms, as hot spots” (Azam

1998). Much higher concentrations of substrates and bacteria can be found in “hot spots” than in the surrounding water column (Pedrós-Alió and Brock, 1983). Pelagic bacteria interact with colloids and particles embedded in the matrix and there is a complex dynamic between DOM and particulate organic matter. These observations go a long way in “fleshing out” the empty space of the ecosystem however, even when we take into account the volume of macro/nano/picoplankton and marine colloids the sum volume of these particles and cells represents less than one hundred thousandth of the total volume of seawater at a cell concentration of 106 ml-1.

Despite the three dimensional matrix and large numbers of particules and cells, the open ocean is an extremely dilute ecosystem composed of mostly ‘empty space’.

1.2.2.1 Nutrient limitation

It is important to consider that the low concentrations of nutrients encountered in oligotrophic environments are in part due to the substrate capture activity of bacteria themselves. Since the open ocean generally lacks an exogenous supply of nutrients, the entire community is ultimately reliant on the release of photoassimilated carbon

10 Chapter 1

from phytoplankton, by diazotrophs in the absence of nitrate/nitrite reductase genes and remineralisation of inorganic nutrients. Taking into account that heterotrophic bacteria dominate the overall biomass of marine ecosystems, by virtue of their surface area and uptake abilities they effectively compete with phytoplankton for mineral nutrients. As a consequence, a close mutual dependence exists between heterotrophic bacteria and phytoplankton (Schut, 1993).

Evidence collected from nutrient addition studies suggest that overall bacterial production is limited by mineral forms of N and P (e.g. Parpais et al., 1996, Sanudo-

Wilhelmy et al., 2002, Thingstad et al., 1998). Specific cases of carbon limitation of heterotrophic bacteria have been found in the equatorial and subarctic Pacific

(Kirchman, 1990, Keil and Kirchman, 1991, Nagata and Kirchman, 1997, Kirchman and Rich, 1997, Church et al., 2000). Aside from methodological problems associated with slow response times of microbial communities (del Giorgio and

Cole, 2000) recent microcosm experiments suggest that the ‘quality’ of added substrates (whether carbon, nitrogen, phosphorus or iron) can have an effect on the efficiency of utilisation of other nutrients, impacting on the observed overall bacterial growth efficiency, productivity and growth rate of communities (del

Giorgio and Cole, 2000). For example, Kirchman (1990) found cases of carbon limitation in the subarctic pacific, however, the best response was found upon addition of amino acids (also a rich source of nitrogen for heterotrophic bacteria,

Keil and Kirchman, 1999). Also, in environments where nitrogen is deficient relative to phosphorus cases have been observed where phytoplankton and the overall bacterial growth rate are limited by phosphorus (Thingstad et al., 1998, Cotner et al.,

1997). It is also possible to design specific probes to assess the physiological state of

11 Chapter 1

microorganisms in situ. For example, Scanlan et al. (1997) generated antibodies for

PstS, a periplamic binding protein required for high affinity uptake of phophate by marine Synechococcus and Prochlorococcus during phosphate depleted conditions.

Such antibodies have been used to demonstrate seasonal phosphate stress of

Synechococcus and Prochlorococcus populations in the Red Sea (Fuller et al., 2004).

Conflicting results of substrate additions and temporal variability make interpretation of the overall picture difficult. A general picture can be based on the principles outlined above with the growth of heterotrophic bacteria limited by the supply of organic carbon and energy, the growth of phytoplankton limited by the availability of inorganic nutrients and a complex dynamic, in time, space and limiting nutrients, existing between different groups. The observations highlight the need for more specific probes to assess the physiological state of bacteria in situ, a task that cannot be achieved without detailed laboratory study of environmentally relevant isolates.

1.3 DISTINCTIONS AND DEFINITIONS

In a manner reflective of the interdependence of heterotrophic and phototrophic bacteria, microbial ecologists and physiologists are mutually dependent upon each other for the generation and testing of valid ecological hypotheses. Unfortunately, the past decades of research into open ocean ecology were heavily based on phenomenological observations rather than rigorous experimentation under controlled conditions. As a consequence, the terminology adopted by both groups can be polemic, phenomenological and profoundly confusing. Terminology can, however, be extremely useful for creating definitions and making distinctions

12 Chapter 1

between life strategies. An explanation of currently used terminology is outlined below in an attempt to highlight useful distinctions.

1.3.1 The size of small organisms

The size of microorganisms varies considerably. Even amongst the prokaryotes

(Archaea and Bacteria) cells with volumes of 0.02 – 180,000,000 μm3 have been isolated. At the smallest end of this size spectrum are the ultramicrobacteria. For years microbial ecologists have recognised that they play important roles in the biological cycling of nutrients and in the formation of biomass. Marine ecologists describe a size-graded series of plankton with dimensions ranging from 0.02 μmto

200 m (Sieburth et al., 1978). The terms femto-, pico-, nano- and micro-plankton are used to distinguish size classes with 10-fold increments between 0.02 μmand

200 μm, respectively (Table 1.1). According to these descriptions, ultramicrobacteria most closely correlate with femtoplankton (or femtobacterioplankton). A range of other terms have been coined to describe small microorganisms (Table 1.1). The prefix which has recently been adopted in a number of fields is “nano”; e.g. nan(n)obacteria, nanobe, nanocell and nanosize. The use of “nano” in this context is intended to refer to a size range much smaller (e.g. tens of nanometers) than in the definition of nanoplankton (0.2-2 μm). The use of the term “nanobacteria” may derive from Morita (1988) where it was described as a synonym for ultramicrobacteria.

The term “ultramicrobacteria” was first adopted by Torella and Morita (1981) to describe extremely small bacteria (less than 0.3 μm diameter) isolated from seawater 13 Chapter 1

that formed ‘ultramicrocolonies’ on agar plates, retained their small cell size when growing on agar plates, and grew very slowly in the presence of high concentrations of nutrients. MacDonnel and Hood (1982) modified this description to include isolates from an estuary obtained by filtration through a 0.2 m membrane and which could form normal-sized colonies on low-nutrient agar that were also observed by Hood and MacDonell (1987). In their review, Schut et al., (1997) further modified the description of ultramicrobacteria to include microorganisms which have a cell volume of less than 0.1 μm3, and which retain this volume irrespective of growth conditions. This description using volume as the defining criteria is particularly useful for studies of natural communities as a range of cell shapes are often encountered, and volume provides a measurement of size that is independent of shape. A list of criteria defining ultramicrobacteria are also described by Velimirov (2001).

1.3.2 Copiotrophs vs Oligotrophs

The use of cell size as a defining criteria is also a source of confusion, particularly because cell size can be variable and there are well documented examples where growth rate, growth stage, nutritional status and physiology can effect bacterial cell size and morphology (e.g. Wase and Patel, 1985, Moyer and Morita, 1989a). Perhaps the best studied example is Escherichia coli where the cell mass varies sixfold over a range of growth rates from 0.6 to 2.5 doublings hr-1 (reviewed in Bremer and Dennis,

1996). The increase in cell mass is reflected by an increases in the total amount of protein, RNA and DNA per cell. Despite this well documented example, two functionally distinct types of cells from the environment that display small volumes

14 Chapter 1

are recognised; ultramicrobacteria and ultramicrocells. Ultramicrocells are formed as a result of starvation induced miniaturisation processes and are characterised by a larger sized (greater than 0.1 μm3) reproductive form and a microcell that can have a volume of less than 0.1 μm3. Unlike ultramicrobacteria which retain a volume of less than 0.1 μm3 when growing, ultramicrocells are dormant, stress-resistant cells, analogous to spores in differentiating bacteria. Accordingly utramicrocells and ultramicrobacteria are products of distinct life strategies adopted by copiotrophic and oligoptrophic bacteria respectively. Copiotrophic bacteria require a high concentration of organic carbon for growth (Poindexter, 1981), they are often associated with nutritionally rich aggregates (DeLong et al., 1993) and are found in relatively higher numbers in coastal regions due to turbulence induced mixing with nutrient rich deeper waters and a considerable input of nutrients from terrestrial sources. When nutrients are scarce starvation induced miniaturisation results in the production of a dormant ultramicrocell (Morita, 1982, 1985, 1988, 1997, Kjelleberg et al., 1993, Holmquist and Kjelleberg, 1993). In contrast, ultramicrobacteria represent the truly low-nutrient adapted oligotrophic species that can become numerically dominant in low nutrient environments (Schut et al., 1997). The presence of these distinct classes of tiny microorganisms is consistent with studies of natural aquatic and soil communities. Microscopic observations of environmental samples reveals cells with volumes of 0.02 – 0.12 m3 (reviewed in Schut et al.,

1997). However, once the environmental samples have been cultured on agar plates, cells typically have volumes on the order of 0.34 - 6.4 m3. Many of the larger cells represent copiotrophs derived from outgrown ultramicrocells, and in practice, are more easily isolated than ultramicrobacteria. Since the isolation of true ultramicrobacteria from marine and soil environments, a key area of research in 15 Chapter 1

contemporary microbial physiology has been to determine what factors affect the culturability of ultramicrobacteria and to compare the physiology of each class.

TABLE 1.1 Description of terms relating to the small size of bacteria

Ultramicrobacteria with a cell volume of less than 0.1 m3 that maintains its size with only minor changes, irrespective of growth conditions. Observed by light microscopy. Ultramicrocells Smaller forms (usually starved) of microorganisms that are larger when actively growing. Usually associated with reductive during starvation. Observed by light microscopy. Nan(n)obacteria Possible synonym for UMB. In the literature usually associated with structures in geological samples with sizes ranging from 0.01 – 0.1 μm. Usually associated with uncultured and unsubstantiated descriptions of microorganisms. Observed by electron microscopy. Femtoplankton 0.02-0.2 μm Picoplankton Marine microorganisms 0.2-2.0 μm Nanoplankton Marine microorganisms 2.0-20 μm Microplankton Marine microorganisms 20-200 μm Other related Dwarf cells/bacteria, lilliputian cells, femtobacterioplankton, terms miniature cells/bacteria, nanocells, nanosized, nanobe, nano- organisms

1.4 OLIGOTROPHIC ISOLATES

1.4.1 Predicted properties of oligotrophs

According to the vital roles of nutrient uptake and utilisation a list of predicted properties were advanced for a model oligotroph at the Dalhelm Conference (Hirsch et al., 1979). The proposed characteristics include: (i) the possession of high surface per volume ratio (cells are expected to be small or possess prostheca), (ii) preferential usage of metabolic energy for nutrient uptake especially during periods of non-growth, (iii) constitutive uptake nutrient ability, (iv) possession of high

16 Chapter 1

affinity, low-specificity transport systems for simultaneous uptake of mixed substrates, (iv) and the establishment of nutrient reserves following nutrient uptake.

The small size of cells would provide a distinct advantage in terms of grazer avoidance (Morita, 1985) and increased efficiency of nutrient uptake, while nutrient uptake mechanisms were expected to have a broad specificity, be inducible and subject to a minimal amount of catabolite repression in order to ensure simultaneous utilisation of the broadest range of substrates (Poindexter, 1979). Oligotrophs were also expected to regulate their biosynthetic rate in line with nutrient uptake rates

(Poindexter, 1979). Finally, oligotrophs were predicted to have the ability to store diverse nutrients in reserves (Hirsch et al., 1979). Since the proposal of these characteristics a range of physiological studies have been conducted to test their validity (reviewed in Schut et al., 1997). Unfortunately very few of these studies were conducted with oligotrophs, highlighting the need to obtain relevant oligotrophic isolates for laboratory studies.

1.4.2 The extinction dilution method for isolating oligotrophs

The difficulty in isolating oligotrophs from the environment is well documented

(reviewed in Schut et al., 1997). Some of the factors which may restrict the ability to isolate and adapt oligotrophs include: 1. intolerance to high concentrations of nutrients, 2. inappropriate growth substrates, 3. the absence of specific vitamins or growth factors, 4. inhibitory growth substrates or other additives, 5. inactivation by the close proximity to other cells (in colonies on agar plates), 6. susceptibility to the oxidative respiratory burst upon upshift and outgrowth in the presence of fresh nutrients and 7. the deleterious effects of lytic phage.

17 Chapter 1

To date the extinction dilution method has been the most successful isolation technique. Extinction dilution has been used to obtain numerically abundant strains of S. alaskensis (Button et al., 1993, Schut et al., 1993) and Cycloclasticus oligotrophus (Button et al., 1993, Wang et al., 1996). In the original description of this procedure seawater samples were diluted in filtered and autoclaved natural seawater without additional nutrients until only a few organisms remained in each dilution tube (Button et al., 1993). Isolation without addition of substrates prevented the possibility of substrate toxicity and removed competition for substrates by less abundant indigenous copiotriophs allowing for long term incubation of potentially pure cultures in the highest dilutions. Long-term incubation of these cultures (6-12 months) in the dark and at 5oC initiated an unknown mechanism that enabled the cells to grow on a rich nutrient medium, i.e. a transition from an obligate to a facultatively oligotrophic state (Schut et al., 1993). The nature of this cell transformation is still unclear, however, during subsequent chemostat experiments one of the isolates, RB2256, became unable to form colonies on nutrient rich agar.

This result suggested that the state of adaptation to the external nutrient concentration was an important factor determining whether an oligotrophic isolate would survive on nutrient rich media or not (Schut et al., 1997b). Sensitivity to nutrient concentration has also been demonstrated for SAR11 isolates that are inhibited by dilute proteose peptone (0.001%) (Rappé et al., 2002). The transition mechanism may involve gradual changes in metabolism or cellular composition that allow cells to survive osmotic stress induced by the initial uptake of nutrients and/or the initial oxidative respiratory burst upon outgrowth.

18 Chapter 1

Despite an incomplete understanding of the mechanisms of adaptation, the extinction dilution method has proven to be reproducible. Insights gained from the study of S. alaskensis RB2256 and C. oligotrophus has highlighted the importance of nutrient concentrations and led to modifications of the extinction dilution method, such as the inclusion of vitamins, antioxidants, and a further reduction in the concentration of complex nutrients. Recent applications of the extinction dilution method have resulted in the isolation of previously uncultivated members of the SAR11 clade

(Rappé et al., 2002) as well as novel from coastal and ocean environments (Cho et al., 2004).

Filtration has also been employed to isolate ultramicrobacteria from the ocean. In contrast to extinction dilution studies however, these attempts have failed (reviewed in Velimirov, 2001). Isolates obtained from the filtrates of 0.2 μm filters have outgrown to larger cells and were therefore likely to be ultramicrocells at the time of filtration. This method however may be useful if the filtrates are initially grown in low-nutrient liquid medium, and then processed in a similar way to those described for the extinction dilution cultures.

19 Chapter 1

TABLE 1.2. General characteristics of selected marine oligotrophic isolates

Marine isolates Characteristics Sphingomonas Isolated as an abundant species from ocean waters near Alaska alaskensis andintheNorthSea Little variation in cell volume (reviewed in facultatively oligotrophic, obligately oligotrophic upon first Cavicchioli et cultivation al., 2003) High affinity, broad specificity nutrient uptake systems predicted to enable successful competition in oligotrophic waters Simultaneous utilisation of mixed substrates Potential to grow at realistic rates in the ocean Absence of a typical starvation stress response Intrinsically resistant to a range of stresses 3.2 Mb Single copy of rRNA operon, maximum 2000 ribosomes cell-1, minimum 200 ribosomes cell-1 Volume = 0.024 m3 Prochlorococcus Most abundant phototrophic organism spp. Global distribution between 40˚N and 40˚S and 0 – 200 m depth Smallest known (Partensky et al., 1.7-2.4 Mb genome 1999) Accounts for more than 50% of and contributes 30- 80% of the total in the oligotrophic oceans Cycloclasticus Isolated from same location as S. alaskensis from Resurrection oligotrophus Bay, Alaska. Dilute cytoplasm (Button et al., ~3 Mb genome 1998) Utilises only a few aromatic hydrocarbons and acetate as growth substrates Kinetic constants for uptake compatible with growth on ambient concentrations of nutrients in seawater volume = 0.01 m3 HTCC1062 Isolated by extinction dilution from Pacific Waters 27 km off (SAR11) Oregon coast Candidatus Member of the ubiquitous SAR11 lineage based on 16S rDNA Pelagibacter gene sequencing ubique 1.54 Mb genome constant cell volume = 0.01 m3 (Rappé et al., maximum cell density in culture is 3.5 x 106 cells ml-1 2002) 0.001% w/v proteose peptone inhibited growth maximum measured growth rate of 0.58 d-1

20 Chapter 1

1.5 DESCRIPTION OF SPHINGOMONAS ALASKENSIS RB2256

S. alaskensis RB2256 is the type strain of Sphingomonas alaskensis (Vancanneyt et al., 2001). It is the one of only a few species of marine oligotrophic ultramicrobacteria cultured to date, and is representative of the dominant, pelagic bacterial species in Ressurection Bay, Alaska. S. alaskensis RB2256 is also recognised as one of the earliest oligotrophic isolates and has therefore been the subject of a number of physiological and molecular studies. This strain possesses a number of characteristics that fit the proposed oligotrophic model (Hirsch et al.,

1979) including: (i) a relatively constant ultramicro-size (< 0.1 m3) irrespective of whether it is growing or starved (Schut, PhD Thesis, 1994, Schut et al., 1997a), (ii) relatively slow maximum specific growth rate ( < 0.2 h-1) (Schut et al., 1993, Schut et al., 1995, Eguchi et al., 1996), (iii) the ability to utilize low concentrations of nutrients through a high affinity, broad specificity uptake system (Schut et al., 1995), and also (iv) the ability to take up mixed substrates simultaneously (Schut et al.,

1995, Schut et al., 1997ab).

1.5.1 Isolation of S. alaskensis

S. alaskensis RB2256 and at least 6 other isolates were obtained from a 10 m depth seawater sample from Resurrection Bay, Alaska, serially diluted in filtered aged seawater medium (FAS) in 1990 (Button et al., 1993, Schut et al., 1993). After a period of months bacterial growth was observed by flow cytometry and epifluorescence microscopy at dilutions of 1 x 106 and5x105 and was dominated (>

50% total cells) by small, short rod-shaped bacteria (Schut et al., 1993). Mixed

21 Chapter 1

cultures with large cells and a variety of different cell types (large rod, cocci, spirilla) were observed to predominate at lower dilutions. The average cell volumes of the small rod-shaped cells were 0.05-0.6 μm3, with the apparent DNA content of 1.0-1.5 fg cell-1. Strain RB2256 and another 8 morphologically similar strains were isolated from serial dilutions demonstrating an initial standing population of 0.2 x 106 cells ml-1 in Resurrection Bay at the time of sampling, indicating that these strains were numerically abundant. These isolates could be subcultured in FAS and oligotrophic medium (2 mg of cassamino acid l-1), but not in high nutrient medium such as full strength marine nutrient agar or broth. Furthermore, growth was never observed on plates, indicating an obligate oligotrophic nature. However, following 6-12 months incubation at 5oC, these isolates gained their ability to form micro colonies on ZoBell

2216E and MPM agars. The acquired ability to grow on richer media represents a transformation from an obligate to facultative oligotroph. In support of the numerical significance of this isolate the presence of this species in Resurrection Bay and the North Sea was also demonstrated by the hybridisation of species specific probes FP1 and FP4 to eight isolates from Ressurection Bay and one isolate (strain

NS1619) from the North Sea (Schut, PhD Thesis, 1994). In addition, the persistence of this species in Resurrection Bay was shown by the detection of cells using

Southern hybridisation of extracted community DNA, more than two years after the initial isolation.

Phylogenetic analysis of strain RB2256 showed that this strain belongs to the -

Proteobacteria lineage, on a single branch within a cluster of Sphingomonas and

Caulobacter (Schut, PhD Thesis, 1994). The genus Sphingomonas was originally described by Yabuuchi et al., in 1990, with the type strain Sphingomonas 22 Chapter 1

paucimobilis but recently split into four new genera by Takeuchi et al. (2001). Prior to 1990, members of the genus Sphingomonas were described as Flavobacterium,

Pseudomonas, Beijerinckia and Arthrobacter. DNA-DNA hybridisation, 16S rDNA sequencing and fatty acid profiles and metabolic characteristics of the remaining isolates indicate that these strains comprise a homogenous genomic species named

Sphingomonas alaskensis with RB2256 as the type strain (Vancanneyt et al., 2001).

At least 7 strains from Resurrection Bay are still available (including RB255,

RB2510, RB2515 and RB2256) from the Culture Collection Laboratorium voor

Microbiologie, University of Gent, Gent, Belgium. To this date, S. alaskensis strains have been isolated from two separate locations: Resurrection Bay Alaska and the

North Sea (Schut et al., 1993), however, the North Sea isolate (strain NS1619) is no longer available.

1.5.2 General growth characteristics

S. alaskensis is an obligately aerobic bacterium that forms opaque yellow, low convex, entire colonies on solid medium (Schut et al., 1993, Vancanneyt et al.,

2001). It has a low specific growth rate (0.13-0.16 h-1 at 23°C) which remains largely unchanged in defined sweater medium with carbon concentrations ranging between 0.8 and 800 mg l-1 (Eguchi et al., 1996). S. alaskensis has the ability to respond rapidly to the addition of excess nutrients. Glucose-limited cultures of S. alaskensis maintain the ability to respond immediately to nutrient up-shift and reach maximum rates of growth without any noticeable lag phase, irrespective of the supplied nutrients (glucose, alanine or acetate) (Eguchi et al., 1996, Fegatella et al.,

1998). The low maximum specific growth rate correlates with a single genome 23 Chapter 1

encoded copy of the ribosomal RNA operon compared to eight or more copies found in the faster growing Vibrio spp. (Fegatella et al., 1998). Even so, S. alaskensis appears to maintain a high cellular concentration of ribosomes and it has been demonstrated that, after periods of outright starvation, a ribosome content 10% of maximum is sufficient to allow cells to immediately respond to nutrient upshift and achieve maximum rates of growth (Fegatella et al., 1998).

S. alaskensis maintains a constitutive high affinity, broad specificity uptake system(s) for amino acids. The alanine uptake system has an affinity for alanine that may exceed any previously reported transport system, is constitutive and is capable of transporting nine other amino acids (Schut et al., 1995). In contrast, the glucose uptake system is inducible and has narrow substrate specificity. Interestingly, alanine and glucose are simultaneously utilised and enable cells to grow at maximum specific growth rates that exceed growth with individual substrates. Both uptake systems use periplasmic binding-proteins which provide an energy dependent transport of substrates against a 105-fold concentration gradient. The uptake systems would enable S. alaskensis to efficiently scavenge and utilise substrates from the environment. Based on the uptake kinetics of these systems, strain RB2256 could grow by using dissolved free amino acids at an in situ doubling time of 12 h to 3 d

(Schut et al., 1995, Schut et al., 1997a), which compares favourably with measured doubling times for bacteria in oligotrophic waters of 5 to 15 d (Fuhrman et al., 1989).

24 Chapter 1

1.5.3 Starvation and stress resistance

In comparison to the well-documented carbon starvation responses of copiotrophic bacteria, the response in S. alaskensis RB2256 differs in many ways (Schut et al.,

1993, Eguchi et al., 1996, Schut et al., 1997, Fegatella et al., 1998, Fegatella and

Cavicchioli, 2000). The most noteable difference is the absence of reductive cell division which has been described for copiotrophs such as Vibrio ANT 300

(Novitsky and Morita, 1976), V. angustum S14 (Holmquist and Kjelleberg, 1993,

Kjelleberg et al., 1993), Aeromonas, Pseudomonas and Alcaligenes (Moriarty and

Bell, 1993). In addition, starvation does not induce increased protection to other the stress inducing agents (known as cross protection), e.g. hydrogen peroxide (25 mM), ethanol (20%), heat (56˚C) (Eguchi et al., 1996) and UV-B in S. alaskensis

(Joux et al., 1999). Instead, S. alaskensis remains inherently resistant to these stresses. For example the viability of growing cultures decreases less than 2-fold after 30 min exposure to 25 mM hydrogen peroxide (Eguchi et al., 1996). Even though S. alaskensis appears to have no phenotypic response to starvation large changes in ribosome levels (Fegatella et al., 1998), rates of macromolecular synthesis and global gene expression (Fegatella and Cavicchioli, 2000) have been observed, indicating that a specific starvation response exists but that it is significantly different to those of copiotrophic bacteria.

The molecular characteristics of the starvation and outgrowth response have been defined in S. alaskensis (Fegatella, PhD Thesis, 2002). The resumption of growth without lag after periods of starvation and the lack of a transient increase in protein synthesis prior to entry into the starvation phase suggest the lack of a typical

25 Chapter 1

stringent response in this organism (e.g. Flårdh et al., 1994, Cashel et al., 1996).

Like other -proteobacteria, S. alaskensis appears to lack RpoS, perhaps the best characterised starvation response regulator in the literature (Hengge-Aronis, 2002).

In the absence of RpoS, cAMP may serve as a global starvation regulator as protein synthesis levels are matched by the intracellular concentration of cAMP throughout the growth phase. The molecular mechanisms controlling gene expression are likely to be complex and novel.

A great deal of information has been gained by direct comparison of S. alaskensis with well studied copiotrophs, which form the basis of the detailed models of gene regulation and phenotypic responses. However, the relevance of studies on a single isolate in batch culture with high concentrations of nutrients need to be addressed.

In the absence of cell density effects such as quorum sensing and oxygen limitation, studies in batch cultures offer a system in which only the physiology of exponential growth in nutrient replete conditions and the transition to nutrient deplete conditions can be studied. Clearly, in order to understand the physiology of an organism that is believed to proliferate by growing slowly on limited amounts of nutrients we need a better system and more geographically and genetically diverse isolates to study.

1.6 HYPOTHESIS

The model oligotroph Sphingomonas alaskensis possesses novel mechanisms for efficient growth and survival during slow, nutrient limited growth and these mechanisms are distinctly different from those possessed by other well-studied marine copiotrophic bacteria.

26 Chapter 1

1.6.1 Overall objectives

The overall objective of this work is to investigate the physiology of novel oligotrophic isolates under environmentally relevant conditions. Since environmental microorganisms are forced to grow at suboptimal rates as a consequence of the limited availability of growth substrates the aim is to examine the physiology of S. alaskensis during nutrient-limited growth. Chemostats provide a useful tool for examining the effects of chronic nutrient limitaion, although traditional chemostat apparatus is cumbersome and experiments consume vast amounts of media. As an aid to high-throughput analysis of replicates and multiple nutrient limitations, small- scale chemostats with very small working volumes are developed.

Despite a wealth of information describing the responses of copiotrophic bacteria to stress and starvation, very little is known about the impact of nutrient limited growth on the physiology of these organisms. For comparative analyses the physiologies of two well-studied organisms, Escherichia coli and a model marine , Vibrio angustum S14 (Nyström and Kjelleberg, 1989, Nyström et al., 1990, 1992,

Holmquist et al., 1993), are also examined under nutrient limitation. Finally, the molecular basis the unique physiology of S. alaskensis will be examined with the aid of proteomics.

1.6.2 Specific aims

27 Chapter 1

1. Chemotaxonomically identify an abundant facultatively-oligotrophic

ultramicrobacterium that was isolated by extinction-dilution culture from the

North Pacific near Toyama Bay, Japan (Chapter 2).

2. Develop small-scale chemostats and methods for analysis of oligotrophic and

copiotrophic bacteria growing under conditions of nutrient limitation

(Chapter 3).

3. Investigate the physiological responses of S. alaskensis to nutrient limited

growth in chemostat cultures and compare the physiology of well-studied

copiotrophic bacteria under the same conditions (Chapter 4).

4. Investigate the effect of nutrient limitation on global gene expression in S.

alaskensis:

a. Establish protocols to generate protein profiles through two-dimensional

gel electrophoresis (2-DE) (Chapter 5).

b. Examine changes in gene expression in carbon limited chemostat cultures:

Perform comparative analysis of cultures growing at different carbon-

limited rates of growth, and identify proteins with altered expression in

each growth condition (Chapter 6).

28 Chapter 2

2MOLECULAR AND CHEMOTAXONOMIC

CHARACTERISATION OF AN ABUNDANT

ULTRAMICROBACTERIUM ISOLATED FROM

THE OLIGOTROPHIC NORTH PACIFIC

29 Chapter 2

2.1 BACKGROUND

A significant focus in marine microbial ecology is the determination of species composition in, as well as the contribution of, each species in biogeochemical cycles.

Molecular studies of microbial diversity and physiological studies with the majority of marine isolates suggest that abundant indigenous marine bacteria are poorly represented in culture collections. Despite advances in assessing species composition and the activity of particular microbial groups in situ, the lack of isolates representing truly low nutrient-adapted microorganisms has limited the understanding of the potential of a large portion of the community.

The extinction dilution method has been used to isolate numerically abundant strains of S. alaskensis and C. oligotrophus from oligotrophic waters in Resurrection Bay,

Alaska (Schut et al., 1993, Button et al., 1993) and representative isolates from the ubiquitous SAR11 and other proteobacterial lineages that have previously defied cultivation efforts (Rappé et al., 2002, Connon and Giovannoni, 2002).

Recently, application of the extinction dilution method by a colleague (Mitsuru

Eguchi) was used to obtain isolates that were abundant in oligotrophic waters near

Japan. After 12 months of serial sub-culturing in filtered autoclaved seawater most of these isolates supported growth in liquid but not on solid medium. However, one isolate, strain AFO1, was able to form colonies on agar plates after 12 months of serial sub culture. Strain AFO1 was isolated from a 10–5 dilution of natural seawater where the microbial count determined by microscopy was 3.1 x 105 cells ml-1

30 Chapter 2

indicating that it was a numerically abundant member of the population at the time of sampling. This chapter presents phylogenetic and chemotaxonomic characterisation of strain AFO1. The new isolate was found to be a genetically distinct strain of S. alaskensis. In light of the increasing number of isolated and characterised

Sphingomonas species and the recent sub division of the genus Sphingomonas into four distinct genera by Takeuchi et al. (2001), the taxonomic characteristics of S. alaskensis RB2256 are re-evaluated.

31 Chapter 2

2.2 MATERIALS AND METHODS

2.2.1 Bacterial strains and culture conditions

S. Alaskensis strains RB2256, RB2510 and RB2515 were obtained from Professor Jan Gottschal. Each strain was revived from stocks that were cryopreserved in 1992, transported on plates, immediately inoculated into artificial seawater medium (Eguchi et al., 1996), incubated overnight and preserved in 10% glycerol at -80˚C. Strain AFO1 was isolated from a depth of 350 m in the North Pacific, 1.5 km from Cape Muroto, Toyama Bay during early summer 1998 by Dr Mitsuru Eguchi. Strain AFO1 was brought onto plates only after 12 months of serial sub-culture in liquid filtered autoclaved seawater (FAS). S. alaskensis strains and strain AFO1 were maintained on ASW medium supplemented with glucose (3 mM), Sphingomonas macrogoltabidus ATCC51380 was maintained on R2A agar (Oxoid), Sphingomonas strain KT-1 (Tabata et al., 1999) was maintained on 3.0 g L-1 yeast extract, 5.0 g L-1 peptone and 0.75% NaCl (YEPS) and subglaciescola ACAM 21 was maintained on Marine Broth 2216 (Difco). For preparation of DNA and fatty acids, S. alaskensis RB2256, strain AFO1 and H. subglaciescola were grown in Marine Broth 2216 (Difco), KT-1 in YEPS and S. marcrogoltabidus in R2Amediumat25°C, and cells harvested at late logarithmic phase (optical density 0.5 at 610 nm).

2.2.2 Alcohol precipitation

DNA was recovered from aqueous solutions by the addition of 0.1 volume of 3 M sodium acetate, pH 5.0, and 2.5 volumes of spectral grade ethanol. Samples were precipitated overnight at -15˚C or for 15 min at 0˚C. The DNA was then pelleted by centrifugation at 14000 x g for 30 min at 4˚C. After washing the pellet with cold 80% ethanol, the DNA was recentrifuged at 4˚C for 2 min, and air dried for 15 min.

2.2.3 Estimation of nucleic acid concentration

Nucleic acid concentrations were estimated spectrophotometrically using a Beckman Du 7500 spectrophotometer at wavelength 260 nm. Double stranded DNA 50 g 32 Chapter 2

ml-1 and single stranded DNA and RNA 40 g ml-1 corresponds to A260 of 1 (Sambrook et al., 1989). The purity of the DNA preparation was determined by the A260/A280 ratio. Solutions with an A260/A280 ratio of 1.8 or greater were considered to be pure. DNA concentrations were also estimated by comparing the relative intensities of DNA samples to standards of known concentration, electrophoresed on agarose gels.

2.2.4 Restriction enzyme digestion

Restriction enzyme digests were carried out in the buffer supplied for that enzyme by the manufacturer, or alternatively in One-Phor-All™ reaction buffer (Pharmacia). Digests contained 1-3 units of enzyme per g of DNA and were incubated for 1-2 h at the temperature recommended by the manufacturer.

2.2.5 Agarose gel electrophoresis and visualisation

DNA preparations such as restriction enzyme digests, PCR amplified DNA and were primarily analysed by electrophoresis on 0.7-1.5% (w/v) agarose (Progen DNA grade) gels. Agarose gel loading dye was added to samples (1:10 v/v) and electrophoresed in TAE buffer containing 0.5 g ml-1 EtBr at 80-120 mA for 1- 2 h (or overnight at 20 mA) on minigels (60 x 60 x 4 mm) or large gels (200 x 200 x 4 mm). Preparative agarose gels prepared aseptically with low melting point agarose (Sea Plaque ® GTG ® Agarose) (1.0 -1.5 % w/v in sterile E-buffer) were used to purify target DNA from contaminants. Gel forming apparatus was disinfected with 95% ethanol. Electrophoresis was performed at 4˚C in sterile TAE-buffer containing 0.5 g ml-1 EtBr. DNA was visualised using a long wave UV lamp. Section 2.5.2.4 outlines the isolation of DNA from preparative gels. DNA and RNA were visualised using a UV transilluminator at wavelength 254 nm and, when permanent records were required, photographs of gels were taken with a Mitsubishi video copy processor P67E.

33 Chapter 2

2.2.6 Genomic DNA extraction

Total genomic DNA was extracted using a modification of the method described by Murray and Thompson (1980). A 2 ml aliquot of late logarithmic phase cells was pelleted by centrifugation, the media decanted, and the pellet resuspended in 567 l of TE. Cells were lysed by the addition of 30 l of 10% w/v SDS and 3.0 l of 20 mg ml-1 proteinase K (Boehringer Mannheim) to give a final concentration of 100 g ml-1 proteinase K and 0.5% w/v SDS. The solution was mixed thoroughly and incubated at 60˚C for 4h before the addition of 100 l of 5 M NaCl and 80 l of 10% w/v CTAB (hexadecyltrimethyl ammonium bromide, Sigma) in 0.7% w/v NaCl. The CTAB/NaCl solution was prepared by slow addition of CTAB (10 g) to 100 ml of 0.7 M NaCl while heating and stirring.

Samples were mixed thoroughly and incubated at 65˚C for 10 min. CTAB complexes were extracted with 1 volume of chloroform:isoamyl alcohol (24:1 v/v) and centrifugation at 12 000 x g for 5 min and the supernatant was transferred to a fresh tube. Any CTAB complexes remaining in the supernatant were extracted with 1 volume of phenol:chloroform: isoamyl alcohol (25:24:1 v/v/v) and centrifugation at 12 000 x g for 5 min. The supernatant was transferred to a fresh tube and nucleic acids were precipitated by the addition of 0.6 volumes of isopropanol. After the tubes were mixed by gentle inversion the nucleic acids were collected by spooling on a glass rod and washed successively in 50%, 70% and 100% v/v ethanol. Spooled and washed DNA was transferred to a fresh tube, dried briefly in vacuuo and resuspended in deionised water. The DNA was subsequently used for 16S rRNA gene amplification, DNA-DNA hybridisation and mol% G+C content analysis.

2.2.7 Amplification of 16S rDNA sequences

Amplification of the almost full-length 16S rRNA gene from AFO1 was performed by PCR using the bacterial consensus 16S rRNA primers 27F and 1494R (Neilan et al., 1997). The amplification mixture (20 l) contained 25-50 ng of DNA, 0.2 mM

(each) deoxynucleoside triphosphates, 0.2 pM (each) primer and 2.5 mM MgCl2. After an initial denaturation step of 5 min at 95˚C, the temperature of the PCR mixture was lowered to 88˚C and 1-2 U Pfu Taq DNA polymerase (Sigma) was 34 Chapter 2

added. Thermal cycling was performed in a PCR Sprint Temperature Cycling System (Hybaid Ltd.) for 25 cycles of: 20 s denaturation at 95˚C, 20 s annealing at 50˚C, and extension at 72˚C for 1.5 min. The final extension step was 5 min at 72˚C followed by storage at 4˚C.

2.2.8 Purification and nucleotide sequencing of amplification products

PCR products were prepared for DNA sequencing by ethanol precipitation. After the addition of 0.1 volume of 3 M sodium acetate and 2 volumes of 80% v/v ethanol the tubes were vortexed for 20 s and incubated at 22˚C for 5 min. The DNA was collected by centrifugation at 12 000 x g for 10 min at 22˚C, the supernatant discarded, and the pellet resuspended in deionised water. DNA sequencing was performed using the PRISM BigDye cycle sequencing system (Applied Biosystems Inc.) and primers 27F, 519R, 530F, 929R, 1114F, 1221R and 1494R (Neilan et al., 1997) according to manufacturers instructions. Sequencing-reaction products were purified by ethanol precipitation, and sequence data was collected from a model 373 sequencer (Applied Biosystems). The DNA sequences were deposited in the GenBank database.

2.2.9 Phylogenetic analysis of 16S rDNA gene sequences

DNA sequences corresponding to Escherichia coli 16S rRNA gene positions 27 to 1433 were aligned using the programs Pileup, ClustalX (Thompson et al., 1997) and PHYLIP (Felsenstein, 1985). The nucleotide alignments were edited manually to resolve positions with ambiguities or gaps. The 16S rDNA distance trees were reconstructed using the neighbor-joining method with Jukes-Cantor corrections (Saitou and Nei., 1987) as implemented by ClustalX. The bootstrap confidence levels for the interior branches of the trees were estimated from 1,000 resamplings of the data (Felsenstein, 1985).

2.2.10 DNA-DNA hybridisation and mol%G+C content analysis

DNA-DNA hybridisation and mol%G+C content analysis were performed on DNA samples extracted at UNSW by Dr John Bowman at the University of Tasmania 35 Chapter 2

using methods adapted from Huss et al. (1983). Briefly, genomic DNA was sheared to an average size of 1 kb using sonication, dialysed overnight at 4°Cin2xSSC buffer (0.3 M NaCl, 0.03 M sodium citrate, pH 7.0), and adjusted in concentration to approximately 60-75 μgml-1. Following denaturation of the DNA samples, hybridisation was performed at the optimal temperature for renaturation (TOR)which was 25°C below the DNA melting temperature and was calculated from the following equation: TOR°C= 48.5 + (0.41 x %G+C). The decline in absorbance over a 40 min interval of DNA mixtures and control DNA samples were used to calculate DNA hybridisation values from the following equation: % DNA hybridisation = (4AB - A - B/2(A x B)) x 100% A and B represent the change in absorbance for two DNA samples being compared and AB represents the change in absorbance for equimolar mixtures of A and B. DNA hybridisation values equal to or below 25% is considered to represent background hybridisation and are thus not considered significant.

The DNA base composition of strains was determined using the spectrophotometric thermal denaturation method described by Sly et al. (1986).

2.2.11 Fatty acid analysis.

Fatty acid analysis was performed on freeze-dried cell pellets by Dr. David Nichols at the University of Tasmania using the following methods. Lipids were extracted using the modified one-phase chloroform:methanol Bligh and Dyer extraction (Bligh and Dyer, 1959). A portion of the total lipid extract was transesterified by reaction at 80°C for 1 h using 3 ml of a methanol:chloroform:hydrochloric acid (10:1:1 v/v/v) solution. After the addition of water, the mixture was extracted with hexane:chloroform (4:1 v/v) to yield fatty acid methyl esters (FAME). Hydroxy functionalities were converted to OTMSi ethers by reaction with bis(trimethylsilyl)trifluoroacetamide (BSTFA) reagent at 80°C for 24 h.

FAME were analysed using a Hewlett Packard 5890 II gas chromatograph and 5970A Mass Selective Detector equipped with a 50 m x 0.22 mm internal diameter cross-linked methyl silicone (0.33 μm film thickness) fused-silica capillary column. 36 Chapter 2

Operating conditions are detailed in Nichols et al. (1986 and 1994). Identification FAME from all samples was achieved by interpretation of component spectra and comparison to those of known standards. Monounsaturated fatty acid double bond position and geometry was determined by GC-MS analysis of their dimethyl disulphide adducts (Nichols et al., 1994).

37 Chapter 2

2.3 RESULTS

2.3.1 Sequencing and phylogeny of strain AFO1

Almost full length (1,414 bp) sequence was obtained for the 16S ribosomal RNA gene (rDNA) of strain AFO1. The 16S rDNA sequence was compared with sequences in the GenBank sequence database and the most closely related sequences were identified by BLAST (http://www.ncbi.nlm.nih.gov/blast). The closest database matches were to two independently determined S. alaskensis RB2256 sequences,

Z73631 (Gottschal et al., submitted to GenBank 1996) displaying 1397/1413 identities, and AF148812 (Button et al., 1998) displaying 1369/1387 identities. High scoring matches were also obtained with S. alaskensis RB2256 co-isolates, strains

RB2515, RB2510 and RB255 (Vancanneyt et al., 2001). In light of the number of ambiguities and unassigned nucleotides in two database entries of S. alaskensis

RB2256 the 16S rDNA gene sequence of this strain was reanalysed using a stock culture cryopreserved in 1992. The 16S rDNA gene was amplified using Pfu high- fidelity DNA polymerase and unambiguous sequence of the rDNA gene was obtained for both DNA strands over a total of 1,414 bp, corresponsing to bases 27-

1441 of the Escherichia coli 16S rRNA sequence. The new sequence supercedes earlier incomplete entries, AF148812 and Z73631, which were obtained using less- stringent methods and error-prone, out-dated sequencing technology and should be used in place of them. The 16S rDNA sequences for strains RB2256 and AFO1 were deposited in the GenBank database under accession number AF378795 and

AF378796 respectively.

38 Chapter 2

The determined sequences for AFO1 and RB2256 along with related sequences from

GenBank were used for constructing phylogenetic trees. Tree topology was similar for parsimony and distance matrix trees and a representative neighbour-joining tree is shown (Fig. 2.1). The marine S. alaskensis strains along with strain AFO1 form a mono-phyletic cluster indicating that strain AFO1 is a strain of S. alaskensis.From the alignments used for tree construction, only one nucleotide change was present between strain AFO1 and RB2256 (1,414 bp) (Table 2.1). S. alaskensis strains

RB2510 and 2515, that were isolated from Resurrection Bay (Vancanneyt et al.,

2001), had one and two nucleotide changes respectively and four alignment gaps compared with the sequence for strain RB2256. In view of the errors for 16S rRNA sequences for strain RB2256 (accession numbers Z73631 and AF148812), these differences may also be sequencing artefacts. The most similar sequence to those in the S. alaskensis cluster is from strain KT-1 (Fig. 2.1, Table 2.1) which has 19 nucleotide differences. The sequence of a Sphingomonas isolated from a hot-spring

(acc. AB015049) was also very similar to AFO1 (98.1% identity, 1311 identities

/1338 residues, no gaps).

39 Chapter 2

TABLE 2.1. Changes and gaps relative to S. alaskensis RB2256T 16S rDNA sequence

Strain Differences1 gaps/insertions1 AF01 (AF378795) 1 0 RB2510 (AF145753) 0 4 RB2515 (AF145754) 1 4 KT-1 (AB022601) 18 1 MBIC3365 (AB015049) 272 02 1 nucleotide changes relative to 1414 bp of RB2256T 16S rDNA sequence (AF378796) 2 nucleotide changes relative to 1338 bp alignment with RB2256T 16S rDNA sequence.

40 Chapter 2

FIGURE. 2.1 Distance-matrix tree of selected Sphingomonas 16S rDNA sequences. DNA sequences corresponding to the E. coli 16S rRNA gene positions 27 to 1433 were aligned using the programs PILEUP and ClustalX (Thompson et al., 1997). Genetic distances were calculated using the method of Jukes and Cantor, and the phylogenetic tree was reconstructed using the neighbor-joining algorithm of Saitou and Nei (1987) as implemented within ClustalX. The phylogenetic tree was plotted using the program nj-plot. The root of the tree was determined using the 16S rDNA gene of Bacillus (BCE277907) as an outgroup. Numbers on branches represent bootstrap values for 1000 repeats (Felsenstein, 1985). The bar indicates 2 nucleotide changes per 100. Accession numbers unless indicated on the figure: (1) AF145754, (2) AF378795, (3) AF378796, (4) AF145753, (5) AB022601, (6) AB015049, (7) AF367204, (8) AF181572, (9) D13723, (10) AF327069, (11) AY081981, (12) D17322, (13) D13727, (14) U20756, (15) X94102, (16) D16144, (17) AF125194, (18) M96746, (19) AF510191 and (20) AF327028.

41 Chapter 2

2.3.2 DNA base composition and DNA-DNA hybridisation

The high degree of sequence identity between strain AFO1 and RB2256 raises the possibility that these two strains are genetically identical. In order to assess the extent of identity between these strains at a high level of resolution the mol% G+C content was determined and DNA-DNA hybridisation was performed. DNA from

Sphingomonas sp. strain KT-1 was used as an external reference. The mol%G+C for

DNA from strains RB2256, AFO1 and KT-1 were 64.8, 65.1 and 65.3, respectively

(Table 2.2). The DNA-DNA reassociation level between genomic DNA from strains

RB2256 and AFO1 averaged 84% while between strains RB2256 and KT-1 DNA hybridisation levels averaged only 34% (Table 2.2). S. macrogoltabidus ATCC

51380 (mol% G+C 65) and H. subglaciescola ACAM 21 were used as control strains

(mol%G+C 59) and both exhibited background renaturation levels of 12 to 23% with strain RB2256. S. macrogoltabidus exhibited significant levels of hybridisation

(44%) with strain KT-1 however below the level indicative of a bacterial species

(Stackebrandt and Goebel, 1994,Wayne et al., 1987). This indicates that strain KT-1 is closely allied to the species S. macrogoltabidus but appears to represent a novel species.

TABLE 2.2 DNA base composition and DNA:DNA hybridisation values between S. alaskensis RB2256, strains AFO1and KT-1, and S. macrogoltabidus.

Strain Mol% G+C (Tm) RB2556 KT-1 %DNA hybridisation S. alaskensis RB2256T 64.8 100 AFO1 65.1 84 KT-1 65.3 34 100 S. macrogoltabidus 65.4 23 44 ATCC51380

42 Chapter 2

2.3.3 Fatty acids

The fatty acid composition of Sphingomonas sp. KT-1 and AFO1 compared to S. alaskensis RB2256 is shown in Table 2.3. S. alaskensis RB2256 and strain AFO1 were grown in ASW supplemented with 3 mM glucose. Sphingomonas strain KT-1 was grown in tap water supplemented with 3.0 g l-1 yeast extract, 5.0 g l-1 peptone,

0.75% w/v NaCl. Samples were taken at mid-exponential growth phase and the washed cell pellets were freeze-dried before fatty acid and methyl ester analysis

(FAME). FAME analysis was carried out by Dr. David Nichols. The two marine strains (AFO1 and S. alaskensis RB2256) possessed a similar fatty acid profile,

(major components being 17:1w6c, 18:1w7c and 17:1w8c) and were distinguished from the freshwater species KT-1 (containing 18:1w7c and 16:1w7c as major components). The fatty acid profiles determined support the phylogenetic relationship between KT-1, AFO1 and S. alaskensis RB2256 as described by 16S rRNA sequence (Fig. 2.1).

43 Chapter 2

TABLE 2.3 FAME analysis of fatty acids from S. alaskensis RB2256 and Sphingomonas strains AFO1 and KT-1.

Percentage composition1 Fatty Acid KT-1 AFO1 RB2256

14:0 0.5 ± tr - - - - 15:0 0.2 ± tr 1.9 ± 0.1 2.0 ±tr 16:0 3.4 ± tr 3.6 ± 0.2 3.2 ±tr 17:0 - - 3.2 ± 0.2 4.0 ±tr 18:0 tr ± tr 0.4 ± tr 0.9 ±1.0 Sum Saturates: 4.1 ± tr 9.1 ± 0.4 10.1 ±1.0

i15:0 - - - - 0.3 ±tr Sum Branched: - - - - 0.3 ±tr

14:1w5c 0.4 ± tr - - - - 15:1w6c - - 0.2 ± tr 0.2 ±tr 16:1w7c 34.2 ± 0.2 4.3 ± 0.2 3.8 ±tr 16:1w5c 3.1 ± tr 0.8 ± tr 0.7 ±tr 17:1w8c 0.3 ± tr 8.1 ± 0.5 10.0 ±0.5 17:1w6c 1.9 ± tr 35.8 ± 1.8 44.7 ±0.3 18:1w7c 42.6 ± tr 24.3 ± 1.2 20.1 ±0.1 18:1w5c 1.2 ± tr 3.8 ± 4.8 0.9 ±tr 19:1w8c - - 0.9 ± tr 1.0 ±0.1 19:1w6c - - 1.2 ± 0.1 1.3 ±0.2 Sum Monounsaturates: 83.6 ± 0.2 79.3 ± 1.0 82.7 ±0.1

2-OH14:0 3.8 ± tr 1.2 ± 0.1 0.6 ±tr 2-OH15:0 - - 5.1 ± 0.3 3.5 ±tr 2-OH16:1w5c 0.3 ± tr 0.4 ± tr 0.1 ±0.1 2-OH16:0 7.6 ± 0.1 3.1 ± 0.2 1.5 ±1.0 2-OH17:1w6c 0.3 ± tr 1.8 ± 0.1 1.2 ±0.2 2-OH18:1w7c 0.4 ± tr - - tr ±tr Sum Hydroxy: 12.4 ± 0.2 11.6 ± 0.6 6.9 ±1.1

Total 100.0 ± tr 100.0 ± tr 100.0 ±tr 1: number of replicates (n) = 3; data presented as average ± standard deviation; tr = trace proportion, defined as less than 0.1%.

44 Chapter 2

2.4 DISCUSSION

2.4.1 Ecological implications of S. alaskensis sp. strain AFO1

The 16S rDNA sequence, mol%G+C, DNA-DNA hybridisation and fatty acid composition indicate that strain AFO1 is almost identical to S. alaskensis RB2256.

The level of sequence identity between strain AFO1 and RB2256 is striking. A result that suggests that these two strains, isolated from the North Pacific near Japan and Alaskan waters respectively, over a period spanning 10 years, are derived from a common ancestor in recent history. Pacific Ocean currents may account for the dispersion of S. alaskensis the 10, 000 km from Alaskan waters to Japan, or vice versa, for example, the North Pacific Intermediate Water current moves water through locations near these two sites (You et al., 2000). The presence of this single species as a numerically significant proportion of the bacteria at geographically remote sampling sites demonstrates the ability of S. alaskensis to proliferate in diverse ocean environments, from the permanently cold waters (4˚C) of Ressurection

Bay (Schut et al., 1993) to the relatively temperate water of the Pacific near Japan.

Strains AFO1 and RB2256 were also isolated from different depths (350 m and 10 m respectively) demonstrating that this species is capable of competing in distinctly different niches within the water column.

In addition to the type strain, S. alaskensis RB2256, and at least six other strains that were isolated from Resurrection Bay in Alaska (Schut et al., 1997a, Vancanneyt et al., 2001), morphologically similar strains have been isolated from the North Sea

(Schut et al., 1997a). A range of distantly related Sphingomonas strains have been isolated from oligotrophic and eutrophic marine environments including the Baltic

45 Chapter 2

Sea (Pinhassi et al., 1997, Pinhassi and Hagstrom, 2000), the North Sea (Eilers et al.,

2000), below sea ice off eastern (Bowman et al., 1997) and as pathogens of corals (Richardson et al., 1998). Estimates of the abundance of Sphingomonas strains in marine environments have been reported by Pinhassi et al. (1997) in the

Baltic sea, (e.g. Bal 46 and BAL35, 18.9% and 2.4% of total bacterial numbers respectively) and by Schut et al. (1993) who isolated Sphingomonas strains with abundances of 15 to 35%.

The fact that two isolates, separated by more than 10,000 km can have almost identical sequences is intriguing, especially when the RB2256 co-isolates, which display less sequence identity at the level of 16S rDNA, are considered. DNA-DNA hybridisation provides some resolution to the relationship between strains AFO1 and

RB2256. The level of genomic DNA-DNA hybridisation (84%) indicates that strain

AFO1 displays significant differences at the genomic level and is therefore a genetically distinct strain of S. alaskensis. Such differences could be the result of genomic rearrangement, deletions, integration of phage or other . In contrast, strains RB255, RB2510 and RB2515 display greater similarity to RB2256 (> 89%) as determined by DNA-DNA hybridisation (Vancanneyt et al.,

2001).

The fatty acid profiles determined support the phylogentic relationship between KT-

1, AFO1 and S. alaskensis as described by the 16S rDNA sequence. Species of all four Sphingomonodaceae genera described by Takeuchi et al. (2001) are characterised by the presence of 2-OH14:0. In addition to this component a series of further 2-OH fatty acids were identified from strains KT-1, AFO1 and RB2256 by

46 Chapter 2

FAME analysis. There is a noticeably higher ratio of 17:1/18:1 and 16:1/16:0 fatty acids in S. alaskensis strains as compared with those from the Sphingomonas,

Novosphingobium and Sphingobium clusters. This general feature is found in other members of the Sphingopyxis genus (Takeuchi et al., 2001) and may serve to distinguish members of this genus from the other genera. It is interesting to note the general quantitative differences for fatty acid groups presented here in comparison with data previously reported for S. alaskensis (Vancanneyt et al., 2001). Together, with the low ratio of 17:1/18:1 and 16:1/16:0 fatty acids measured for the phylogenetically closely related freshwater species, strain KT-1 (1394/1414 identities in 16S rDNA sequence, 98.6%, Table 2.3), these differences highlight the dependence of fatty acid profiles on growth conditions, e.g. high NaCl vs no NaCl, media richness and growth phase.

2.4.2 Reclassification of Sphingomonas alaskensis to Sphingopyxis alaskensis

Phylogenetic analysis of existing Sphingomonas 16S rDNA gene sequences indicate that the currently known members of the genus could be resolved into four distinct clusters (Takeuchi et al., 2001). In support of these, distinctions based on gene sequencing, chemotaxonomic and phenotypic differences were noted between each cluster. Three new genera were proposed in addition to the genus Sphingomonas sensu stricto. According to this reclassification scheme S. alaskensis strains RB2256,

AFO1 and related strains belong to the genus Sphingopyxis with Sphingopyxis macrogoltabidus IFO 15033T (formerly Sphingomonas macrogoltabidus) as type strain. Characteristics that were used for distinguishing Sphingopyxis from the other

‘Sphingomonas’ clusters include greater than 97% 16S rDNA identity to the type

47 Chapter 2

strain, the absence of nitrate reduction, the accumulation of the polyamine spermidine rather than homospermidine (Takeuchi et al., 2001), and the fatty acid profiles as outlined in the results and the discussion below. It is notable that individual members of each cluster have been isolated from diverse environments and that the 16S rDNA delineation does not support groupings based on the from which they were isolated. This is reflected by the diverse range of environmental conditions from which members of the Sphingomonodaceae have been isolated (Takeuchi et al., 2001 and references therein). Unusual environments from which Sphingomonads have been isolated include the accretion ice above

Vostok (Christner et al., 2001), the deep subsurface (Balkwill et al., 1997), activated sludge (Neef et al., 1999) and hydrocarbon polluted soils (Baraniecki et al., 2002). It is interesting to note the high degree of 16S rDNA sequence similarity of strains

RB2256 and AFO1 to strains KT-1 and MBIC3365 (Table 2.1) considering that KT-

1 was isolated from natural mineral water and strain MBIC3365 was isolated from a halophilic spa. As evidenced by their wide distributions and metabolic diversity members of the family Sphingomonodaceae appear to be generalists, capable of adapting to a wide range of physical and biochemical environments.

2.4.3 Conclusion

Strain AFO1 is almost identical to S. alaskensis RB2256. Work done by colleagues has established that morphology, genome size and stress resistance profiles of strain

AFO1 are identical to S. alaskensis RB2256. This collective work was published by

Eguchi et al. in 2001. The isolation of AFO1 indicates a broad geographical distribution, persistence and environmental significance for S. alaskensis genotypes.

48 Chapter 2

In light of the overall level of similarity between AFO1 and the type strain RB2256, only RB2256 was used for the remainder of studies performed in the following chapters.

49 Chapter 3

3CONTINUOUS CULTURE METHODS FOR

INVESTIGATING PHYSIOLOGICAL

RESPONSES TO NUTRIENT LIMITATION IN

MARINE BACTERIA

50 Chapter 3

3.1 BACKGROUND

In order to comprehend microbiological processes and the environmental processes in which they participate, it is essential to understand the physiology of microorganisms with respect to the prevailing environmental conditions. In the oligotrophic ocean the limited availability of nutrients limits microbial growth and has a major impact on microbial physiology. Growth limitation relates largely to the availability of utilisable nutrients, such as carbon, nitrogen and phosphorus, however, the availability of trace metals, vitamins and a variety of physicochemical factors may also impact on the abundance of microorganisms.

Organisms in oligotrophic environments inhabit a constant limitation. Although simple laboratory systems cannot capture the richness and complexity of natural ecosystems, such systems do permit careful measurement of cell physiology under well-defined and reproducible conditions. Furthermore, well-defined laboratory systems offer a starting point for studying responses to controlled environmental changes and allow the rigorous experimental testing of hypotheses (e.g. Gottschal,

1990a, Velicer et al., 1999). The traditional and perhaps most straight-forward method of microbial cultivation is the batch culture. Within this system initially all nutrients are in excess allowing the organism to grow at an optimal, unlimited rate.

At a point where the concentration of one or more nutrients become limiting the exponential growth rate diminishes and eventually ceases and the organism enters a starvation state. The main disadvantage of this system is that exponential growth and near zero growth in stationary phase are the growth modes that may be studied in detail. The intervening phases of decelerating growth, corresponding to the period of

51 Chapter 3

nutrient limitation, rapidly merge into one another and are therefore difficult to study in isolation.

While still poorly studied, there is expanding interest in the effects of growth rate, as may be imposed by nutrient limitation in chemostats, on the physiology of microorganisms. For example, cell size, cellular composition and starvation survival are affected by growth rate in Vibrio sp. strain ANT-300 (Moyer and Morita, 1989a,

1989b), and slow growth regulates porin expression and induces cAMP- and RpoS- dependent gene expression in E. coli (Matin and Matin, 1982, Schultz et al., 1988,

Notley and Ferenci, 1995, 1996, Liu and Ferenci, 1998, Tweeddale et al., 1998).

Other regulatory mechanisms that respond to the level of nutrient limitation in E. coli, such as the stringent response and regulation of mutation rates themselves, have been identified (Cashel et al., 1996, Notley-McRobb et al., 1997, Ferenci, 1999).

In response to starvation copiotrophic bacteria become resistant to a variety of stresses (e.g. Matin, 1990, Nyström et al., 1990ab, Nelson et al., 1997).

The chemostat is a basic piece of laboratory apparatus that allows the study of bacteria growing at sub-optimal rates. Chemostat studies have begun to occupy an increasingly central role in ecological and evolutionary studies (e.g. Velicer et al.,

1999, Notley-McRobb and Ferenci, 1999, Stahl et al., 2004). The environment created by a chemostat is one of the few completely controlled experimental systems for testing microbial growth and competition. The theory and procedures for continuous cultivation have been reviewed extensively (Gottschal, 1990a and

1990b). A basic diagram of a chemostat set-up, comparing the two types of chemostats employed in this study is presented in Figure 3.1. The basic parameters

52 Chapter 3

defining chemostat growth are presented in Table 3.1. In the chemostat, the cell density and culture volume are kept constant and the growth rate of the culture is fixed externally by the rate at which fresh medium is supplied. After a period of equilibration the specific growth rate () is considered to be equal to the dilution rate

(D):

=D (h-1) in which: D =F/V F=(volume of fresh medium fed to the chemostat per hour) V=(volume of the chemostat culture)

In this way a specific growth rate, which is indicative of a specific level of nutrient limitation, can be imposed upon a culture. Cellular density within the culture can be set by the concentration of the limiting nutrient in the medium . In other words, the growth rate and level of nutrient limitation can be set independently of the culture density (over certain physiological limits). Steady state is achieved when the growth rate of cells is exactly balanced by the rate of removal of cells by washout.

By adjusting the supply rate of fresh medium, the growth rate of the bacteria may be varied from almost zero to close to maximal growth rate. Therefore, the growth rate of a steady-state chemostat culture is directly proportional to the extent of nutrient limitation imposed.

The physiological adaptation of bacteria to starvation is strongly dependent on the preceding growth conditions (Gottschal, 1990, Moyer and Morita, 1989a, 1989b).

With S. alaskensis early chemostat experiments revealed low plate viabilities (<

0.001%) for cells that were starved after glucose limited growth in continuous

53 Chapter 3

cultures (Schut, PhD Thesis, 1994). In contrast, cells that were starved following growth in batch culture were initially 100% viable on agar plates and cells from the dilution tubes from which the strain was initially isolated were fully viable after one year at 5˚C. The reasons for this discrepancy are not clear but thought to relate to specific metabolic adaptations required for nutrient uptake during substrate limited growth. Loss of plate viability may also relate to a switch from a facultative to an obligately oligotrophic state similar to that displayed upon initial isolation.

Although the basic procedures for chemostat cultivation of S. alaskensis have been established (Schut, PhD Thesis, 1994), the basic parameters of substrate limitation, growth yield and plate viability examined in the present section in a search for clues as to the oligotrophic/facultative oligotrophic conversion and to establish methods for a more detailed characterisation of physiology and molecular biology of the organism under nutrient limited conditions. Small scale, low cost chemostats (50 ml working volume) were designed to facilitate high-throughput and genetic experiments of S. alaskensis. The function of these apparatus were tested under working conditions and found to satisfy the theoretical considerations of chemostats.

The impact of nutrient limitation on the morphology and growth yield of S. alaskensis is presented and initial comparison is made with a model marine copiotrophic isolate, Vibrio angustum S14 grown under similar conditions.

54 Chapter 3

A

Figure 3.1. Schematic diagram of the chemostat apparatus used in this study. (A) Large scale (500 ml working volume) conventional chemostat setup with separate air and medium inlets and (B) small scale (50 ml working volume) chemostat with combined sterile air and medium inlet directly into the culture liquid.

55 Chapter 3

Table 3.1 Assumed physiological constants and their relationships during steady state cultivation in chemostat culture

Constants Volume of reactor (V) Flow rate of fresh media (F) Cell numbers Specific growth rate ( [h-1]) = Dilution rate (D [h-1]) [External nutrients] And D (h-1)=F/V=(h-1)

3.1.1 Small scale chemostats

In standard chemostat apparatus the inflow is kept separate from the culture, in order to prevent growth back along the lines, and enters the reactor from above, usually sheathed in a flow of sterile air. Chemostat theory assumes that a chemostat reactor is a reproducible homogenous, well-mixed, equilibrium, steady state. However, at low rates of inflow these assumptions may be violated, as the drop-wise addition of fresh media results in a heterogeneous, pulsed-batch system rather than an homogenous well-mixed system. With standard equipment the minimum experimentally determined drop size was found to be ~ 10 l. At a flow rate of 500

l h-1, equivalent to a dilution rate of 0.01 h-1 in a 50 ml chemostat, there would be

50 drops h-1 or one drop every 80 seconds. The amount of mixing is also a critical factor as bacteria immediately below the drop experience a transient higher concentration of nutrients than bacteria distant from the drop.

56 Chapter 3

To alleviate this problem a different system was employed where the flow of fresh media, controlled by a peristaltic pump, passed through a small aperture (~ 0.2 mm dia., provided by a 29 1/2 gauge syringe needle). Sterile pre-humidified air was supplied over the aperture at a rate of between 5 and 10 ml min-1 resulting in vaporisation of the media, in other words the production of small aerosol droplets, from the end of the narrow gauge needle. The air/medium mixture is fed directly into the culture fluid through a 0.5 mm dia. stainless steel tube. The flow of air is used to provide constant flow of fresh media and oxygen to the reactor while preventing grow-back along the feed line. The flow of air was provided at a sufficient rate in order to allow a maximum respiration rate but not result in evaporative loss of volume from the reactor.

3.1.2 Theoretical considerations: growth rates and nutrient flux

Oligotrophic environments are defined by an overall low concentration of nutrients as well as by a low nutrient flux (<1 mg C l-1d-1, Poindexter, 1981). Even so, the growth rate of natural microbial communities in oligotrophic open ocean ecosystems may be greater than 1 doubling per day. The maximum measured growth rate of S. alaskensis is 0.22 – 0.24 h-1 when grown in complex media containing 8 g l-1 yeast extract at 37˚C (Eguchi et al., 1996, Figure 3.3) but less than 0.02 h-1 when grown at

4˚C (unpublished results).

Running chemostats at low dilution rates and low temperatures is time consuming and impractical. The chosen growth temperature of 30˚C represents a compromise between the temperature of waters from which S. alaskensis strains have been

57 Chapter 3

isolated (4˚C to 8˚C) and the optimal growth temperature of the organism in artificial sea water media. Despite high cell densities used in these studies (108 ~109 cells ml-

1 ,OD600 ~ 0.5) the flux of carbon per cell per day falls below the value used to define an oligotrophic environment. At a specific growth rate close to max in chemostat culture, the calculated flux of carbon per 105 cells is 0.025 mg d-1, while at = 0.027 h-1 the flux of carbon per 105 cells is 0.0034 mg d-1. In other words, the growth rates studied in this work, which ranged from 0.02 h-1 to 0.205 h-1 (max in ASW for S. alaskensis), reflect a nutrient flux per cell per day similar to those encountered in oligotrophic environments.

58 Chapter 3

3.2 MATERIALS AND METHODS

3.2.1 Bacterial strains, media and culturing conditions

Sphingopyxis alaskensis RB strains were obtained from Prof. Jan Gottschal (University of Groningen, The Netherlands). For cultivation in liquid batch cultures, S. alaskensis strains and V. angustum S14 were grown in an artificial seawater medium ASW (Eguchi et al., 1996) supplemented with 3.0 mM D-glucose (ASWG) unless otherwise stated. General maintenance cultures of S. alaskensis and V. angustum S14 were carried out in Marine Nutrient Broth 2216 (MNB, Difco).

For chemostat cultivation the media used was ASW with modifications outlined below. Glucose-limited feed medium contained glucose at a concentration of 3.0 mM. For ammonium-limited feed medium the concentration of NH4Cl was adjusted to 0.94 mM and the D-glucose concentration was 5 mM. For mixed amino acid- limited feed medium the concentration of casamino acids was 0.05 % (w/v) and glucose was omitted. The pH of media for batch and chemostat cultures was maintained at 7.8 by the addition of morpholinepropanesulfonic acid (MOPS) buffer (1.0 g l-1) or by continuous automatic adjustment with sterile NaOH (0.25 M). The pH of all media was adjusted to 7.8 prior to autoclaving. Batch cultures were grown at 30°C, with orbital shaking at 100-150 rpm. Chemostat cultivation was carried out in glass culture vessels (450 ml working volume) magnetically stirred at 400 rpm and maintained at a constant temperature of 30°C.

Small-scale chemostat cultivation was carried out in purpose-built 100 ml glass erlenmeyer-shaped culture vessels, with small headspace volume, equipped with two stainless steel baffles and stirred magnetically at 400-500 rpm. The temperature was maintained by continuous flow of water through a water-jacket from a temperature regulated water bath. Medium entered the vessel through a stainless steel needle at the bottom of the vessel (dia. 0.5 mm). Filter sterilised, pre-humidified air was supplied through the same aperture. The specific growth rates imposed on chemostat cultures were alternated between high and low values. Experimental chemostat and

59 Chapter 3

batch cultures were inoculated from ASW pre-cultures established from cryogenically preserved stocks. In order to minimise the effects of mutant selection in chemostat cultures, steady-state was assumed after growth between five to seven generations. Growth of batch and chemostat cultures was monitored by measuring the optical density at 433 nm and by plate counts on ASWG and MNB solidified with 12.5 g l-1 bacteriological agar (Bacto). Culture purity was assessed regularly by plating and phase-contrast microscopy of wet mounts at 100 x magnification under oil immersion using an Olympus EHA (Tokyo, Japan) phase contrast microscope.

3.2.2 Total cell counts and morphology

Total cell counts were determined by 4’6-diamidino-2-phenylindole (DAPI) nuclear staining (Schut et al., 1993). Briefly, samples (1.0 ml) were withdrawn from a culture, fixed in 0.2% glutaraldehyde for 1 min and then permeabilised with 0.1% v/v Triton X-100 to improve stain penetration. Cells were then stained for 15 min with DAPI (0.5 g ml-1, Sigma) and collected onto a black 0.1 m pore size polycarbonate membrane filter (Poretics, USA), and rinsed with 4 ml of a prefiltered 3% w/v NaCl solution. The filter was mounted on a glass side under a cover slip and observed with an Olympus epifluorescence microscope with a 365 nm optical bandpass filter. At least 300 cells and 10 fields were counted for each sample.

Cell dimensions, length (l) and width (w), were estimated on fixed and unfixed, stained and unstained cells with the aid of an eyepiece micrometer mounted on an Olympus EHA phase contrast microscope or an Olympus epifluorescence microscope. Absolute cell dimensions (l and w) of S. alaskensis RB2256 were determined from electron micrographs with the aid of Carnoy2.0 software (www.carnoy.org). Electron micrographs were obtained by transmission electron microscopy of fixed and unfixed cells, negatively stained with Uranyl-acetate using a Hitachi H-7000. Cell volume was calculated from length and width measurements using the formula: volume = (/4)w2(l-w/3).

60 Chapter 3

3.2.3 Viability measurements

Viable counts of S. alaskensis RB2256 and V. angustum S14 were estimated from the number of CFU on Marine Nutrient Agar (MNB, Difco Marine Agar 2216), and ASW-glucose (ASWG) solid media consisting of ASW, 3 mM D-glucose. Dilution series were carried out in ASW buffered with MOPS (1.0 g l-1,pH7.8).Viable counts of E. coli strains were estimated from the number of CFU on nutrient agar. Dilution series were carried out in 0.9% w/v NaCl. Colonies on drop plates (Hoben, 1982) were counted with the aid of a binocular microscope (25 x magnification) after 3 and 6 days of incubation in the dark at 30°C for S. alaskensis, after overnight incubation at 30°CforV. angustum S14 and after overnight incubation at 37˚C for E. coli strains. At least 5 spots from duplicate plates were counted for each experiment.

61 Chapter 3

3.3 RESULTS

3.3.1 Validation of mini chemostats

To test the continuity of flow and mixing in the constructed miniature chemostats the following experiment was set up: bacterial culture was replaced with some dilute acid and phenolphthalein or mixed indicator in a vessel with 50 ml working volume, the solution being fed in was milliQ H2O made alkaline with a few drops of 1.0 N

NaOH and also contained indicator (phenolphthalein or mixed indicator). In much the same way as observed in an acid/base titration pink drops could be seen at the fresh media inlet as liquid was pumped into the vessel. When the vessel was stirred with a magnetic stirbar at 200-400 rpm the pink cloud dispersed and gradually faded to a colourless fluid. Only when the mixing was complete would the colour fade.

This setup was trialed with various liquid inflow rates from 0.5 to 30.0 ml h-1

(equivalent to Dilutiuon rates of 0.001 and 0.6 h-1) and air inflow rates between 1.0 and 20.0 ml min-1. Over the entire range of liquid flow rates mixing was observed to be instantaneous, i.e. no colour was observed at the medium inlet at all (Table 3.2).

An air inflow rate of between 4.0 and 6.0 ml min-1 was sufficient to ensure a constant flow of liquid off the end of the injection needle while minimising the overall turbulence in the vessel, with the exception at high rates of liquid inflow (> 20 ml h-

1), where an air inflow of 10 ml min-1 was used. The experiment was repeated at similar dilution rates with a large scale chemostat (450 ml working volume, F = 0.45 to 270 ml h-1). In the large scale chemostat visible mixing times ranged between 3 and 5 s, with longer mixing times observed at lower dilution rates where the medium entered the vessel as distinct drops separated by as much as 5 min.

62 Chapter 3

TABLE 3.2. Mixing time, a measure of homogeneity in ‘classic’ large scale and small scale chemostat culture vessels. Mixing time was measured as the time taken for any trace of colour to disappear from the vessel

Test vessel mixing time no mixing > 60 s Classic chemostat (1,000 ml) 3-5 s mini chemostat (50 ml) < 1 s

3.3.2 Substrate limitation

To determine the range of concentrations for which D-glucose was the limiting factor in ASW medium S. alaskensis was grown in ASW batch culture supplemented with various concentrations of D-glucose (0.0-9.0 mM). Final cell density was estimated by measuring the optical density (433 nm) of cultures at the onset and up to 12 h after the onset of stationary phase. A linear relationship was observed between D- glucose concentration and growth yield between 0.0 and 4.5 mM D-glucose. This result indicates that D-glucose was the growth-limiting nutrient for S. alaskensis cultures grown in liquid ASW medium when supplied at concentrations up to 4.5 mM. These results show that other factors in the media become limiting above 4.5 mM even though a higher growth yield was obtained when D-glucose was supplied at 6.0 mM. In a similar manner to D-glucose, ammonium (NH4Cl) was defined as the limiting component in ASW when its concentration was altered to 0.094 mM in the presence of 4.5 mM D-glucose. Mixed amino acids (cassamino acids, Sigma

Chemical Co.) were defined as the growth-limiting nutrient when supplied at 0.05%

63 Chapter 3

w/v in the absence of glucose. Trehalose was defined as the limiting component when supplied at 2 mM in the absence of glucose.

3.3.3 Steady-state growth and viability of S. alaskensis and V. angustum S14 in

chemostat culture.

In contrast to the low plate viability observed by Schut (1994), steady-state S. alaskensis cultures grown at different specific growth rates, from =0.020 h-1 to

=0.20 h-1, in glucose limited chemostats and at max (0.205 h-1) in batch culture exhibited no significant differences in final optical density, direct counts of DAPI stained cells and CFU ml-1 on ASWG and MNB solid media (Fig 3.2). Thus, under the growth conditions used throughout this study, a viable but non-culturable

(VBNC) state (Kell et al., 1998) was not observed for S. alaskensis. Molar growth yield is defined as the amount of biomass produced per mole of substrate consumed.

The molar growth yield of S. alaskensis when grown in glucose limited chemostats was 29.3 ± 1.8 g of protein per mole of glucose consumed. This value is similar to values reported previously (Schut, PhD Thesis, 1994). Additionally, no significant differences in molar growth yield were observed for cultures grown at any specific growth rate (Fig 3.3).

In contrast to S. alaskensis, the steady-state growth yield and CFU ml-1 of V. angustum S14 grown aerobically in ASWG chemostats depended on the growth rate of the culture (Fig 3.2). A maximum steady state growth-yield of 13.6 ± 2.4 g of protein mole-1 was obtained for rates of growth above =0.195 h-1 and were constant and equivalent to the growth yield and cell numbers obtained in batch culture (12.7 ±

64 Chapter 3

-1 8 -1 3.1 g protein mole ,OD600 0.45, ~9.0 x 10 CFU ml ). At rates of growth below

=0.195 h-1 the molar growth yield and CFU ml-1 steadily decreased to 1.1 ± 0.6 g protein mole-1 (6.9 x 107 CFU ml-1) at =0.023 h-1. This change represents approximately a 10-fold reduction in growth yield and cell numbers at a low rate of growth (=0.023 h-1) compared to a higher rate of growth (=0.195 h-1). No significant difference between direct counts of DAPI stained cells and CFU ml-1 on solid media were observed at any growth rate and no differences in viability were observed between CFU ml-1 on MNB and ASWG.

65 Chapter 3

1.0E+10 -1 Cell Count and CFU ml

1.0E+09 0 50 100 150 200 250 Time (h)

FIGURE 3.2 Development of total cell numbers and CFU in a glucose limited chemostat of S. alaskensis . At t =48 h after inoculation the pump was switched on

(D = 0.026 h-1). The development of total cell numbers was monitored by cell counts of DAPI stained cells () and CFU ml-1 on ASWG ()andMNB(). For all time points the standard error was less than 20 %.

66 Chapter 3

1.00E+10 50 )

1.00E+09 40 -1

1.00E+08 30 -1

CFUml 1.00E+07 20

1.00E+06 10 Molar growth yield (g mol

1.00E+05 0 0 0.2 0.4 0.6 0.8 Growth Rate (h-1)

FIGURE 3.3 Molar growth yield and cell numbers of S. alaskensis and V. angustum S14 grown in glucose limited chemostats at different specific rates of growth. Growth yield of S. alaskensis ()andV. angustum S14 () was estimated from the amount of protein produced per mole of substrate consumed and total cell numbers of S. alaskensis ()andV. angustum S14 () were estimated from CFU ml-1 on ASWG and MNB. Cultures were grown aerobically in ASW supplemented with 3 mM glucose at 30°C.

3.3.4 Morphology and cell dimensions

S. alaskensis was grown at various temperatures in a range of growth media in batch and under nutrient-limiting conditions in chemostat and cellular morphology was examined by phase contrast and transmission electron microscopy. Under all growth conditions examined cells were regular rods, however chains or filaments up to 70

m in length, with septa separating individual cells, were observed. Under conditions 67 Chapter 3

of glucose limitation at various growth rates single cells predominate but cells growing in chains, usually two, three or four cells long were always present. At higher dilution rates the chains become more frequent and longer. When grown at

10˚C in batch culture without shaking cells are almost entirely singular. When grown under nitrogen limitation in chemostat culture, chains of 2, 3 and 4 were also present but the cells tended to floc together.

The size of single cells growing in ASW during log phase at 30˚C was estimated with an eyepiece micrometer using phase contrast microscopy (100 x magnification).

Cell length ranged from 0.75 to 3.0 m. Cell width was much less than 0.5 m and estimated to be 0.2 to 0.3m. The size differences within these estimates arise from the natural variation in cell length that is associated with cell division. Septa were clearly visible in the middle of most of the longer single cells suggesting that these cells were in the process of dividing. Estimates of cell dimensions were limited by the resolution of the phase contrast microscope where the graduations on the micrometer correspond to 0.5 m at 100 x magnification. The observed changes in cell size and volume under different culture conditions were too subtle to be quantified with the light microscope

In order to obtain a more precise estimate of cellular dimensions S. alaskensis cells were observed using transmission electron microscopy. Glutaraldehyde-fixed negatively stained cells taken from exponentially growing and starved batch cultures and slowly growing steady state chemostat cultures were observed at 50,000 x magnification. Cell dimensions and volumes were calculated from the measurements taken from TEM and phase contrast images (Tables 3.3 and 3.4). In a defined

68 Chapter 3

medium (ASW), the cell volume of growing cells was ~0.05 μm3 and reduced approximately 2-fold after 24 h starvation. In a complex medium (MNB) cell volumes were approximately twice those observed in the defined medium.

69 Chapter 3

TABLE 3.3 Size and volume measurements for S. alaskensis determined by transmission electron Microscopy and phase contrast microscopy.

Growth conditions Length (μm)1 Width (μm)1 Volume (μm3) ASWG 30˚C Exponential 1.12 ± 0.2 0.31 ± 0.1 0.077 = 0.026 h-1 0.91 ± 0.15 0.28 ± 0.1 0.050 Starved 24 h 0.90 ± 0.15 0.28 ± 0.1 0.050 MNB (Complex) Exponential 1.20 ± 0.2 0.33 ± 0.1 0.093

1: measurements are the average for at least 20 cells

TABLE 3.4 Size and volume measurements for V. angustum S14 determined from estimates based on phase contrast microscopy.

Growth Length (μm)1 Width (μm)1 Volume conditions (μm3) ASW 30˚C Exponential 4.1 ±0.4 0.6 ± 0.2 1.1 = 0.195 h-1 2.2 ± 0.3 0.45 ± 0.15 0.28 = 0.026 h-1 1.2 ± 0.35 0.44 ± 0.15 0.16 Starved 24 h 0.94 ± 0.2 0.41 ± 0.15 0.11 Complex MNB 4.5 ± 0.4 0.8 ± 0.25 2.1 Exponential

1: measurements are the average for at least 20 cells

70 Chapter 3

3.4 DISCUSSION

3.4.1 The impact of nutrient limitation on the viability, morphology and cell

dimensions of S. alaskensis and V. angustum S14

In order to examine gross physiological responses to the level of nutrient limitation we examined growth yield, cell numbers and viability of S. alaskensis and V. angustum S14 at different rates of growth in chemostat culture. No differences in viability were observed on rich and poor media and CFU correlated with direct counts obtained by DAPI staining. These results indicate that the viability of S. alaskensis and V. angustum S14 was not affected by nutrient limitation and these organisms did not enter a viable but non culturable state throughout the range of growth rates used in this study. Results obtained here are in contrast to the loss of plate viability observed by Schut (1994), following continuous culture of S. alaskensis in glucose limited chemostats. The differences may relate to the composition the media used (e.g. the inclusion of MOPS buffer) or the relatively short window of generations (5 – 7) used to define steady state in this study. For the remainder of physiological experiments carried out in this thesis viability was always determined on complex (MNB) and defined (ASW) nutrient media and any differences were noted.

Nutrient limitation in chemostat culture was shown to have an impact on the cell size of S. alaskensis and V. angustum S14. Consistent with earlier observations (Schut,

PhD Thesis, 1994, Eguchi et al., 1996), the cell size of S. alaskensis varied marginally (Table 3.3). Even under starvation conditions, the reduction in cell size

71 Chapter 3

was limited to about 2-fold. Even though the errors associated with the resolving power of the light microscope and fixation of cells for TEM are relatively large, changes in cell size were clearly evident. The changes in cell size observed in this study agree with measurements obtained by collaborators in Japan using atomic force microscopy for S. alaskensis during starvation (Mitsuru Eguchi, personal communication). These data indicate that the cell volume of this bacterium may change depending on growth medium and growth phase, however, the cells retain a volume of less than 0.1 μm3. In direct contrast, the cell volume of V. angustum S14 varied by more than 10-fold while growing in glucose limited chemostats (Table

3.4). The largest difference in cell volume (20-fold) was recorded between cells growing exponentially in MNB (2.6 m3) and cells that had been starved for 24 h in

ASW following growth in chemostat at =0.05 h-1 (0.11 m3). The volume estimates obtained for V. angustum S14 are similar to those obtained for a closely related

Vibrio sp. ANT 300 grown in batch culture and over a range of dilution rates in chemostat culture (Moyer and Morita, 1989a). The size estimates obtained for Vibrio sp. ANT 300 range from 5.94 m3 when growing in batch culture down to 0.046 m3 when starved after growth in a chemostat at a dilution rate of 0.015 h-1. Despite a decrease in cell size with decreasing growth rate, reductive cell division was always observed for Vibrio sp. ANT300 in response to starvation to the extent that cell volume was reduced by between 69 and 95%. It was also shown that the preceding growth conditions had an impact on starvation survival with nutrient limited Vibrio sp. ANT300 cultures were better adapted to long periods of starvation.

An important observation is that both of these copiotrophic species exhibit distinctly oligotrophic properties when growing in chemostat cultures. That is, they continue to 72 Chapter 3

grow, even at very low rates of growth, rather than choose to become growth arrested cells. These observations suggest that copiotrophic species may be adapted to a variety of conditions rather than just a ‘feast or famine’ existence. It is tempting to speculate if there may be a threshold level of nutrients at which copiotrophic bacteria make a commitment to become growth arrested cells, as appears to be the case with

Escherichia coli growing in chemostat cultures (e.g. Notley-McRobb and Ferenci,

1996, Ihssen and Egli, 2004) or if slowly growing chemostat cultures are composed of a heterogeneous population of growing and non-growing cells (e.g. Moyer and

Morita, 1989b, also reviewed in Koch, 1997).

Ultramicro-size is a feature shared by all S. alaskensis species (Vancanneyt et al.,

2001). The size and morphology of S. alaskensis is similar to indigenous bacteria observed in natural populations by fluorescence microscopy. Size estimates made by transmission electron microscopy on S. alaskensis indicate that it is one of the smallest free-living and replicating pure cultures of cells known.

The fact that UMB are often associated with oligotrophic conditions is consistent with these small sized cells having a high surface area to volume ratio, which enhances the opportunity for cells to uptake nutrients from the environment. By containing less mass, UMB also require less nutrients than larger cells to produce progeny. UMB may also be less subject to grazing pressure by larger predators (e.g. marine protozoa). These properties are consistent with a UMB that is adapted to growth under nutrient limitation and not simply a dormant member of the population.

These physiological properties are corroborated by ecological data demonstrating the presence of S. alaskensis as one of the most numerically abundant microorganisms

73 Chapter 3

from a number of ocean sites (Schut, PhD Thesis, 1994, Eguchi et al., 2001, Chapter

2).

3.4.2 Growth yield and maintenance energy of S. alaskensis and V. angustum

S14 in response to nutrient limitation

The level of nutrient limitation in chemostats was found to have no effect on the growth yield of S. alaskensis, however nutrient limitation had some influence on the yield of V. angustum S14 (Fig 3.3). Maintenance energy demand (MED) is defined as the amount of energy consumed without a corresponding increase in biomass

(Pirt, 1965). Studies of bacterial growth at low dilution rates have indicated a need for maintenance energy to explain the proportional reduction in growth yields at low growth rates (Pirt, 1965, 1982). For example, Chesbro et al. (1979) describe a 50% increase in MED for E. coli as the growth rate of the organism is reduced below approximately 0.05 h-1 in a recycling fermentor. To explain this it was postulated that below a threshold rate of growth significant amounts of the available energy was diverted to maintenance and survival functions rather than biomass production (Pirt,

1965). Maintenance and survival processes could include functions such as DNA repair, protein turnover, maintenance of osmotic pressure and membrane potential.

The effect of growth rate was also examined on growth yield. A comparison was performed between S. alaskensis and V. angustum S14 to determine whether S. alaskensis had a superior ability to convert substrates into biomass and cell numbers at low growth rates. In glucose-limited chemostats, the growth yield of S. alaskensis was consistent between growth rates from 0.02 h-1 to maximum growth rates (0.20 h-

74 Chapter 3

1) (Fig. 3.3); in fact, growth yields tended to increase slightly as growth rates were lowered. In contrast to S. alaskensis, the growth yield of V. angustum S14 decreased to approximately 10% of maximum levels at the lowest growth rates tested (0.02 h-1).

Growth yields started to decrease when growth rates fell below approximately 25% of maximum growth rates (Fig. 3.3). Above a growth rate of 0.20 h-1, growth yields for V. angustum S14 were constant and equivalent to the growth yield obtained in batch cultures. The data for CFU was confirmed by optical density (433nm) and direct counts of DAPI stained cells for both S. alaskensis and V. angustum S14.

These results may be explained in terms of MED. These data indicates that maintenance energy remains the same in S. alaskensis, even at a low growth rates

(equivalent to 10% of max and above), whereas in V. angustum S14 it increases below a growth rate of 0.20 h-1. This result is in line with earlier observations of the continuous reduction of cell size and concurrent increase in stationary phase survival for another marine Vibrio (sp. ANT300) with decreasing growth rate in chemsotats

(Moyer and Morita, 1989a). Notley and Ferenci (1996) observed that expression levels of RpoS and RpoS-dependent enzymes were as high in glucose limited E. coli cultures at = 0.1– 0.2 h-1 as they were in starved cultures and attributed the high- level induction of RpoS-dependent protective functions to the increase in maintenance energy requirements. Furthermore, the increase in the abundance and influence of stationary phase sigma factors at low growth rates in E. coli and B. subtilis suggest significant overlap between the patterns of gene expression during nutrient limitation with the patterns of gene expression ascribed to stationary phase or carbon starvation responses (Notley and Ferenci, 1996, Tiech et al., 1999,

Schweder et al., 1999). This explanation may also be extended to V. angustum S14

75 Chapter 3

since similar starvation induced survival mechanisms have been described but as yet not correlated with the level of limitation (Srinivasan and Kjelleberg, 1998). These results suggest that S. alaskensis and V. angustum S14 adopt different strategies for dealing with extreme nutrient limitation.

The present data suggest that oligotrophic bacteria are efficient at utilising available nutrients, thereby enabling them to produce more cells from the same substrate pool as compared to other bacteria. These observations support the concept of a growth- rate trade-off between the oligotrophic and copiotrophic life-strategies. Further evidence in support of this comes from examination of different growth patterns of two facultative copiotrophic marine isolates in response to different modes of nutrient addition and competition (Pernthaler et al., 2001) results from this study support the trade-off hypothesis and suggest that there may be a continuum of life strategies, from oligotrophic to copiotrophic, adopted by marine microorganisms.

Such a hypothesis could be examined further by direct competition studies in nutrient limited cultures or alternatively by pure culture studies conducted in situ in dilution chambers.

76 Chapter 4

4PHYSIOLOGICAL RESPONSES TO NUTRIENT

LIMITATION AND STRESS FOR AN

OLIGOTROPHIC ULTRAMICROBACTERIUM,

SHINGOPYXIS ALASKENSIS

77 Chapter 4

4.1 BACKGROUND

The induction of stress resistance mechanisms is a defining characteristic of non- spore forming bacteria that are able to survive long periods of starvation. The ability to resist the damaging effects of oxidative stress, such as hydrogen peroxide, is an ecologically relevant characteristic due to endogenous and exogenous oxidative stress being a common challenge to microorganisms in aquatic environments

(Cooper and Zika, 1983, Arana et al., 1992, Gourmelon et al., 1994, Heikes et al.,

1996, Oda et al., 1997, Schut, et al., 1997, Nyström, 1999). Reactive oxygen species can cause damage to DNA, RNA, protein and lipids and as a consequence, cells have evolved a broad range of mechanisms to cope with this type of stress (reviewed in

Storz and Imlay, 1999). A significant number of starvation induced proteins are geared toward the protection and renaturation of proteins suggesting that the main role of oxidative stress defence proteins is the protection of cellular machinery from endogenously derived oxidative stress caused by ongoing metabolism (Nyström,

1999). In this context, defence against oxidative stress may also be a significant issue for slowly growing cells, since their capacity to replace damaged cellular components by new synthesis is limited by availability. Additionally, despite low in situ growth rates bacterial communities in oligotrophic environments display relatively high endogenous respiration rates, identifying the respiratory transport chain as a potentially significant source of harmful reactive oxygen species.

Ultraviolet radiation may also have an impact on marine microorganisms. As a consequence of diminishing ozone levels the flux of UV-B radiation at the sea surface is increasing. The depth at which the irradiance of mid UV (320 nm) drops to

78 Chapter 4

1% of surface levels can be up to 49 m in clear open ocean waters indicating that UV stress is not restricted to the sea surface microlayer (Moran and Zepp, 2000). Based on its biological consequences, UV can be divided up into UV-A (315 - 400nm) and

UV-B (280 ~ 315 nm). UV-B can have a major impact on microorganisms, mainly through direct damage to DNA, while both UV-A and UV-B can cause indirect damage to biological molecules through reactive oxygen intermediates. In addition to diverse oxidative defence systems a range of starvation induced DNA protection and repair mechanisms have been characterised in diverse bacterial species that provide protection against UV radiation (Nystrom, 2003).

In contrast to our understanding of the physiological responses to starvation, little is known about the physiological changes that occur when nutrients are present, but at concentrations that result in sub maximal rates of growth, such as those that are likely to occur in the ocean. One of the characteristics that distinguishes S. alaskensis from a range of copiotrophic bacteria, such as V. angustum S14, is the lack of phenotypic changes and induced cross-protection in response to starvation (Eguchi et al., 1996). Although starvation-induced cross protection is not observed, S. alaskensis maintains a high level of resistance to a variety of stress inducing agents

(hydrogen peroxide, heat, ethanol and UV) regardless of whether it is growing or starved (Eguchi et al., 1996, Joux et al., 1999).

Previous studies have shown that S. alaskensis was more resistant to hydrogen peroxide, heat and ethanol when grown in glucose-limited chemostats at a dilution rate of 0.027 h-1 (approximately 15% of the max) than logarithmic phase or starved cells from batch culture (Eguchi et al., 1996). In view of the fact that slow, nutrient

79 Chapter 4

limited growth is likely to be the growth condition most often encountered by oligotrophic bacteria, we reasoned that the high degree of resistance observed in chemostat grown cells may be triggered by nutrient limited growth, rather than starvation.

In order to determine the types of mechanisms and regulatory processes that S. alaskensis has evolved to cope with nutrient limitation and stress, in this study the physiological responses of cells grown in nutrient-limited chemostats at a range of fixed sub-optimal growth rates is examined. Growth in chemostats permitted continued growth at a fixed rate, while also permitting the use of different limiting nutrients (e.g. carbon or nitrogen). The impact of nutrient limitation on hydrogen peroxide and UV stress resistances was examined in detail. In addition to UV and hydrogen peroxide responses to a range of other natural or simulated stresses can be used as important indicators of the physiological state of microbial cells. In this study the examination of heat, freeze-thaw and ethanol stresses are extended to chemostat cultures. In order to directly compare the responses of a S. alaskensis with those of typical copiotrophic bacteria, assessment of the impact of nutrient limitation on the control of hydrogen peroxide resistance was extended to include Escherichia coli and V. angustum S14. Another aim of this present study was to identify the pigment component of S. alaskensis, believed to be a carotenoid, and to determine if the cellular pigment played a role in the observed patterns of resistance to various stresses.

80 Chapter 4

4.2 MATERIALS AND METHODS

4.2.1 Bacterial strains, maintenance and culturing conditions

Escherichia coli MC4100 was obtained from Prof. Tom Ferenci at the University of Sydney, Australia. Escherichia coli was cultivated in liquid Minimal Medium A (MMA, Miller 1992) and the growth temperature was 37˚C. MMA was supplemented with 0.5 mg l-1 thiamine and 1.1 and 11.0 mM glucose for chemostat and batch culture respectively. Chemostat culture of E. coli was carried out as described above. Luria Bertani (LB) medium contained 10 g l-1 tryptone, 5 g l-1 yeast extract and 10 g l-1 NaCl. Overnight bacterial liquid cultures were inoculated with a single colony from the stock plate via a sterile toothpick or a scraping from a glycerol stock (0.85 ml culture and 0.15 ml 80% glycerol, snap frozen in dry ice/ethanol bath and stored at -80˚C), and grown at 37˚C with vigorous shaking (220 rpm). Plating was performed aseptically by streaking a single colony from an old to fresh agar plate with a flamed loop, and incubation at 37˚C overnight. Ampicillin (100 g ml-1), kanamycin (30 g ml-1), tetracycline (12.5 g ml-1), chloramphenicol (15 or 30 g ml-1) and streptomycin (500 g ml-1) were added to liquid media and plates as required.

4.2.2 Viability measurements

Viable counts of S. alaskensis RB2256 and V. angustum S14 were estimated from the number of CFU on Marine Nutrient Agar (MNB, Difco Marine Agar 2216), and ASW-glucose (ASWG) solid media consisting of ASW, 3 mM D-glucose. Dilution series were carried out in ASW buffered with MOPS (1.0 g l-1,pH7.8).Viable counts of E. coli strains were estimated from the number of CFU on nutrient agar. Dilution series were carried out in 0.9% w/v NaCl. Colonies on drop plates were counted with the aid of a binocular microscope (25 x magnification) after 3 and 6 days of incubation in the dark at 30°C for S. alaskensis RB2256, after overnight incubation at 30°CforV. angustum S14 and after overnight incubation at 37˚C for E. coli strains. At least 5 spots from duplicate plates were counted for each experiment.

81 Chapter 4

4.2.3 Stress exposure experiments

Cells were withdrawn from mid-exponential batch cultures (OD433 0.4 for S. alaskensis RB2256 and V. angustum S14 or OD580 0.3 for E. coli strains) or steady- state chemostat cultures and exposed to (i) Hydrogen peroxide (2-75 mM) in which 2.0 ml of cells was pippetted into a 5 ml glass test tube containing a 50–100 x concentrated solution of hydrogen peroxide, freshly prepared from a 30% stock solution, for the specified times up to 90 min at 30˚C. (ii) Heat stress as experienced by a shift from 30˚C to the stress temperature (52-58˚C) in which 200 l of culture was transferred to a thin walled PCR tube and then placed in the heating block of a Hybaid gradient thermal cycler for up to 60 min. (iii) Freeze-thaw as experienced by a shift from 30˚C to –20˚C in which 200 l of culture transferred to a 1.5 ml microfuge tube was immediately transferred to –20˚C for 24 h, thawed for 30 min at room temperature and then diluted and plated. (iv) Ultraviolet radiation in which 100 l of culture was spread on a glass slide and exposed to UV in an ultraviolet RPN 2500 cross-linker (Amersham Bioscience). Stress experiments were performed at least twice on independent cultures and viability was determined according to the previous section. The survival fraction for stress treated samples was calculated for each sample separately as a percentage of the survival of untreated control samples.

4.2.4 Determination of catalase activity

Catalase activity was determined with a Clark oxygen electrode by the method of Rørth and Jensen (1967). The concentration of oxygen in oxygen-saturated deionised water at 25°C was assumed to be 253 M (Wilhelm, 1977). The amount of -1 enzyme activity that decomposed 1 mol of H2O2 to 0.5 mol O2 min at 25°Cwas defined as 1 unit of activity. Total catalase activity was determined by subtracting background respiration and oxygen production due to the spontaneous of H2O2. Comparisons of the catalase activity of whole bacteria were made after the addition of H2O2 (final concentration 1 mM) to 2.0 ml of culture (3.0 x109 cells ml-1). Enzyme activity was calculated from the initial reaction rate from at least five replicates and at least two independently prepared samples.

82 Chapter 4

Additionally, catalase activity was measured spectrophotometrically according to the method of Beers and Sizer (1952). Briefly, catalase assays were carried out by monitoring the disappearance of H2O2 at 240 nm using the molar absorption co- -1 efficient of 43.48 l mol for H2O2. Comparisons of the catalase activity at different rates of growth were made using 100 l of crude cell lysate. A standardised solution of H2O2 in ASW (A240 0.55) at 25˚C was used for these assays. Crude cell lysates were prepared by sonication on ice for 2-5 min with a Branson Sonifier on a 30% duty cycle and an output setting of 3.5. All assays were repeated to give at least 20 rate determinations for the first minute of reaction and catalase activity was recorded -1 -1 as the rate of decomposition of H2O2 min (mg protein) . Protein concentration was determined using the Bicinchoninic Acid (BCA, Smith et al., 1985) kit (Sigma) with BSA as a standard.

83 Chapter 4

4.3 RESULTS

4.3.1 The effect of growth rate on hydrogen peroxide resistance in S.

alaskensis.

To determine whether growth rate affected the stress resistance of S. alaskensis, cells were grown at different dilution rates in a glucose-limited chemostat at 30°C, samples were removed and survival was monitored after exposure to 25 mM hydrogen peroxide for 60 min. The specific growth rates used ranged from 0.020 to

0.18 h-1; corresponding to 10 to 87% of the maximum specific growth rate (0.21 h-1).

In addition, the resistance was determined for cells grown at maximum rate in batch culture.

Two distinct levels of resistance to hydrogen peroxide stress were observed (Fig.

4.1). When cultures were grown at rates above a threshold value of 0.13 h-1, the level of stress resistance was equivalent to that of batch grown cells. This level of resistance was the same for cultures grown at five different rates of growth between

0.13 and 0.18 h-1. Between a growth rate of 0.13 and 0.02 h-1, S. alaskensis cultures were 10,000 times more resistant to hydrogen peroxide stress than faster growing cells, maintaining a viability of more than 35% for at least 60 min of exposure to hydrogen peroxide.

For every experiment that was performed at a low dilution rate, an equivalent experiment was subsequently performed at a high dilution rate (and vice versa). This demonstrated that the history of the growth of the cultures had no bearing on the

84 Chapter 4

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01 % Survival

1.0E-02

1.0E-03

1.0E-04 0 0.05 0.1 0.15 0.2

Growth Rate (h-1)

FIGURE 4.1 Percent survival of S. alaskensis grown at different specific growth rates in glucose-limited chemostats after exposure to 25mM hydrogen peroxide for 60 min. For the two trendlines the data were grouped: (), low specific growth rates (=0.020, 0.03, 0.08, 0.10, 0.11, 0.12 and 0.13 h-1) and (), high specific growth rates (=0.14, 0.15, 0.16, 0.170 and 0.18 h-1). Survival of logarithmic phase cells in batch culture () is also included. Duplicate samples were taken from chemostats at every specific growth rate that was tested. Six individual chemostat runs were tested and at least one sample taken from a high and a low specific growth rate. Colony forming units were scored using the drop-plate method on MNB and ASWG. No differences in survival were observed on ASWG and MNB for each time point. A standard deviation of less than 15% was observed for each time point.

85 Chapter 4

observed levels of stress resistance. Furthermore, it showed that if mutants arose in the chemostat, as it is unlikely that the same mutant would arise under both regimes of shifting the growth rate up and shifting the growth rate down, mutants could not account for the observed growth rate dependency of stress resistance.

To determine whether the kinetics of survival for cultures from fast and slow rates of growth was similar, survival in 25 mM hydrogen peroxide was examined throughout a 60 min time course for glucose-limited chemostat cultures grown at a range of specific growth rates from 0.02 to 0.18 h-1, and in batch culture grown at a rate of

0.21 h-1 (Fig. 4.2). These results are consistent with a bimodal response to hydrogen peroxide stress, resulting in two distinct physiological states of S. alaskensis, whereby highest levels of resistance are achieved by growth under glucose limitation at specific growth rates below 0.13 h-1.

4.3.2 The effect of temperature on peroxide resistance

In order to determine if resistance to hydrogen peroxide was regulated by growth rate per se, temperature was also utilised to limit the growth rate of cultures. The maximum growth rate of S. alaskensis grown in ASWG at 10˚C was 0.017 ± 0.002 h-

1. As shown in Figure 4.3 S. alaskensis cultures cultivated at 10˚C in ASWG were remarkably resistant to hydrogen peroxide, exhibiting no loss in viability after exposure to 50 mM hydrogen peroxide for 1 h at 10˚C.

To determine whether the temperature at which cells were exposed to hydrogen peroxide had an effect on survival a number of experiments were conducted. No loss

86 Chapter 4

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01 % Survival

1.0E-02

1.0E-03

1.0E-04 0 20406080100 time (min)

FIGURE 4.2 Percent survival of nutrient limited chemostat grown S. alaskensis cells following exposure to 25 mM hydrogen peroxide for up to 60 min. Experiments were performed twice and colony forming units counted using the drop-plate method on MNB and ASWG Samples were taken directly from steady state glucose-limited chemostats at =0.02 h-1 (), =0.08 h-1 (), =0.13 h-1 (), =0.16 h-1 (), =0.18 h-1 () and in batch, max=0.21 h-1().

87 Chapter 4

of viability was observed when cells taken from a batch culture grown at 10˚C were transferred to 30˚C for up to 1 h (without hydrogen peroxide). However, when the same cells were exposed to 25 mM peroxide at 30˚C the survival fraction dropped to less than 0.0001 % CFU ml-1 within 20 min. Conversely, cells that were exponentially growing at 30˚C and exposed to peroxide at 10˚C for up to 1 h exhibited no loss in viability. These results demonstrate that the survival of S. alaskensis following exposure to hydrogen peroxide was dependent upon temperature at which the cells were exposed. Despite the difficulty in dissecting out the compounding effects of thermal stress suffered during a shift from 10˚C to 30˚C the peroxide resistance of batch cultures grown at 10˚C does not support that hydrogen peroxide resistance is regulated by growth rate per se in the presence of excess nutrients

4.3.3 The effect of carbon vs nitrogen limitation on peroxide resistance

To determine whether the link between growth rate and hydrogen peroxide resistance was affected by the nature of the limiting substrate in the chemostat, the responses of carbon-limited cultures (Fig. 4.2 and 4.4b) were compared to nitrogen-limited cultures (Fig. 4.4a). Ammonium-limited cultures were grown at low (0.05 h-1)and high (0.15 h-1) rates of growth and samples, taken from cultures after a steady state had been obtained, were subjected to 25 mM hydrogen peroxide. These nitrogen- limited cultures were far more sensitive to hydrogen peroxide than any of the glucose-limited cultures, displaying almost undetectable survival after 20 min exposure. Futhermore, no difference in stress resistance was observed at these two specific growth rates.

88 Chapter 4

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01

1.0E-02 % Survival

1.0E-03

1.0E-04

1.0E-05 0 20406080

H2O2 Concentration (mM)

FIGURE 4.3 Percent survival of S. alaskensis in response to the concentration of hydrogen peroxide. Experiments were performed twice and colony forming units counted using the drop-plate method on MNB and ASWG. () Batch culture at 10˚C, () batch culture at 30˚C and ()=0.02h-1 at 30 ˚C

89 Chapter 4

To further test whether the observed bimodal response under glucose-limitation was due to carbon/energy-limitation per se, and not a specific effect of glucose metabolism, glucose was replaced with mixed-amino acids. When mixed-amino acids were used as the sole carbon and energy source, the pattern of hydrogen peroxide resistance across low (0.05 h-1) and high (0.15 h-1) rates of growth (Fig.

4.4b) was essentially the same as for glucose-limited cultures (Fig. 4.2). These data indicate that the effect of growth rate on hydrogen peroxide resistance is a general phenomenon in S. alaskensis linked to growth under carbon and/or energy limitation.

4.3.4 The effect of growth rate on hydrogen peroxide resistance in V.

angustum S14.

To determine whether the rate of growth affected the ability of V. angustum S14 to survive hydrogen peroxide, in a similar way as S. alaskensis, cells were grown in glucose-limited chemostats at 0.02, 0.2 and 0.60 h-1, followed by exposure to stress in 2 mM hydrogen peroxide for up to 60 min (Fig. 4.5). In ASW defined minimal medium with 3 mM glucose as the sole carbon and energy source, the growth rates

0.02 h-1 and0.60h-1 correspond to 4% and 97% of the maximum specific growth rate

(0.62 h-1), respectively. The resistance of these chemostat grown cells was also compared with cells grown at their maximum rate in batch culture and following 24 h starvation (Fig. 4.5).

Starved cells were much more resistant (4.2% survival after 60 min) than cells grown at max (1.7 x 10-5% after 30 min), while glucose-limited chemostat grown cells exhibitedanintermediatelevelofresistance(lessthan3x10-4% after 60 min). In

90 Chapter 4

1.0E+03

1.0E+02 A

1.0E+01

1.0E+00

1.0E-01

1.0E-02

1.0E-03

Percent Survival 1.0E-04

1.0E-05

1.0E-06 0204060 1.0E+03 B 1.0E+02

1.0E+01

1.0E+00

1.0E-01

Percent Survival 1.0E-02

1.0E-03

1.0E-04 0204060 Time (min)

FIGURE 4.4 Percent survival of nutrient limited chemostat grown S. alaskensis cells following exposure to 25 mM hydrogen peroxide for up to 60 min. Experiments were performed twice and colony forming units counted using the drop-plate method on MNB and ASWG. Results of representative experiments are shown. A standard deviation of less than 20% was observed for each time point. A. Samples were taken directly from ammonium-limited chemostats at =0.15 h-1 () and =0.05 h-1 (). B. Samples were taken directly from mixed amino acid-limited steady state chemostats at =0.15 h-1 () and =0.05 h-1 ().

91 Chapter 4

contrast to the bimodal response exhibited by S. alaskensis, the resistance of glucose limited V. angustum S14 varied only by about 10-fold and, most interestingly, no correlation with growth rate could be identified. For example, after 60 min exposure to hydrogen peroxide, the percentage survival was 2.5 x 10-4,1.2x10-5 and 3.4 x 10-

4 for growth rates of 0.023 h-1, 0.195 h-1 and 0.602 h-1, respectively.

4.3.5 The effect of growth rate on hydrogen peroxide resistance in E. coli

E. coli K12 (UNSW002900) cells were grown in MMA in glucose-limited chemostats at 0.10, 0.11, 0.51 h-1, at max (estimate) in batch culture or starved for

24 h following unrestricted growth in batch culture and subsequently exposed to 25 mM hydrogen peroxide for up to 60 min (Fig. 4.6). Of the growth conditions tested, starved cells were most resistant (18% viable after 60 min) and cells from log-phase batch culture (max = 0.8 h-1) were the least resistant (<0.01% after 60 min). Cells from glucose limited chemostats showed intermediate levels of resistance (0.1-1% after 60 min) and no clear correlation between growth rate and survival was identified. The results shown in Fig. 4.6 are results of at least two independent experiments. A similar pattern of resistance was observed when the hydrogen peroxide concentration was 75 mM instead of 25 mM however the rate of death was higher (results not shown). These data indicate that the general response of the cells to hydrogen peroxide stress, varies depending mainly on the mode of growth (batch vs chemostat) for E. coli K12. Similar experiments have been conducted with another E. coli strain (MC4100) (L. Notley-McRobb and T. Ferenci, unpublished).

Peroxide stress resistance in strain MC4100 was significantly higher than for strain

K12 under all growth conditions tested (75 mM vs 25 mM). The overall pattern of

92 Chapter 4

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01

1.0E-02 Percent Survival

1.0E-03

1.0E-04

1.0E-05 0204060 Time (min)

FIGURE 4.5 Percent survival of V. angustum S14 following exposure to 2mM hydrogen peroxide for up to 60 min after growth at various rates in glucose limited chemostats. Samples were taken directly from steady state chemostats at =0.02 h-1 (), =0.2 h-1 (), =0.60 h-1 (), and from batch cultures in logarithmic phase () and after 24 h starvation (). Experiments were performed twice and colony forming units counted using the drop-plate method on MNB and ASWG for each time point. For all data points the standard deviation was less than 15%.

93 Chapter 4

resistance in strain MC4100 was similar to strain K12, with starved cells being the most resistant (69% viable after 25 min), however, cells taken from chemostat at a high dilution rate (0.6 h-1), rather than log-phase in batch, were the most sensitive

(1% after 10 min). Log-phase and low growth rate (0.1 h-1) cells exhibited intermediate levels of resistance (5%-25% after 25 min). Taken together these results describe a general trend in hydrogen peroxide stress resistance in E. coli strains that is growth mode dependent and a complex function of growth rate.

4.3.6 Catalase activity and hydrogen peroxide resistance in S. alaskensis

When hydrogen peroxide was added to liquid cultures of S. alaskensis cells, small bubbles developed between 5 to 30 minutes. This occurred irrespective of growth rate and whether the cells were grown in batch or chemostat. The production of bubbles may result from the production of O2 due to decomposition of hydrogen peroxide by catalase.

To quantitatively determine whether catalase activity correlated with the degree of hydrogen peroxide resistance observed, catalase activity was measured in whole cell suspensions of S. alaskensis grown at high (=0.15 h-1) and low (=0.08 h-1) rates of growth in glucose limited chemostats and at max in batch culture.

The catalase activity varied marginally between cells grown at low (6.6 U mg protein-1) and high (4.2 U mg protein-1) specific growth rates in chemostat and in batch culture (1.1 U mg protein-1) (Table 4.1). In particular, the less than 2-fold

94 Chapter 4

102

101

100 % Survival 10-1

10-2

10-3

10-4 0 10 20 30 40 50 60 Time (min)

FIGURE 4.6 Percent survival of log-phase, starved and chemostat grown E. coli K12 UNSW002900 after exposure to 25 mM hydrogen peroxide for up to 60 min. Samples were taken directly from log-phase batch culture (), after 24 h of starvation () and from steady state glucose limited chemostats at =0.51 h-1 (), =0.11 h-1 (), =0.10 h-1 () and exposed to 25 mM hydrogen peroxide for up to 60 min.

95 Chapter 4

difference in catalase activity between cells grown at low and high rates of growth did not correlate with the 10,000-fold difference in stress resistance (Fig. 4.1).

Total peroxidase activity was also measured spectrophotometrically by monitoring the disappearance of hydrogen peroxide at 240 nm (not shown). A substance present in the culture supernatant and crude cell extract with a large absorbance peak at 230 nm interfered with measurement of total peroxidase activity by this method.

Variation between replicates was high, however, the measured peroxidase activity was not significantly different to catalase activity and was less than 10.0 U mg protein-1. These data clearly indicate that catalase and peroxidase activity does not account independently for the observed bimodal response to hydrogen peroxide in S. alaskensis and suggest that other cellular factors are responsible.

96 Chapter 4

TABLE 4.1. Catalase activity measured in whole cells of S. alaskensis grown in glucose-limited chemostats and batch cultures. One unit of catalase activity is defined as the amount of enzyme required to degrade one mol H2O2 to 0.5 mol O2 in one minute.

Growth Catalasea activity Catalasea activity conditions U mg protein-1 U108 cells-1 Log phase 1.10 0.0117 =0.08 h-1 6.63 0.0704 =0.15 h-1 4.16 0.0442 a standard deviation, 8 - 12%

4.3.7 The effect of growth rate on resistance to ultraviolet radiation in S.

alaskensis

The survival of exponentially growing cultures of S. alaskensis was monitored following exposure of up to 10,000 J m-2 of broad spectrum UV radiation (UVR).

The survival of exposed S. alaskensis decreased exponentially with increasing doses and no distinct differences in CFU were observed on MNB and ASWG (Fig. 4.7).

These results are similar to the results obtained by Joux and co-workers (1999). By way of example, in this study a survival fraction of 0.1 % was recorded after a dose of 4,000 J m-2 compared with a survival fraction of 1.0% following exposure to

3,000 J m-2 of UV-B alone reported by Joux et al. (1999). Taken together, these results suggest that the UV-A component of broad spectrum UVR used in this study has little effect, either positive or negative, on the survival of S. alaskensis in comparison with UV-B.

97 Chapter 4

To examine the effect of glucose limitation on the UV resistance of S. alaskensis samples were taken directly from steady state chemostats growing at =0.03, 0.06,

0.08, 0.16 and 0.17 h-1 and from batch culture at max (0.21 h-1) and exposed to

4,000 Jm-2 of broad spectrum UV radiation. The pattern of survival of glucose- limited, chemostat-grown S. alaskensis was different to the bimodal survival exhibited in response to hydrogen peroxide. The survival fraction following exposure varied by about 2 – fold from 0.001 to 0.1%. A very weak trend in UVR resistance vs specific growth rate was observed, with bacteria grown at higher rates being more resistant (Fig. 4.8). No significant differences in viability were recorded when the stressed cells were plated on MNB or ASWG.

4.3.8 The effect of growth rate on heat stress and freeze-thaw resistance in S.

alaskensis

Heat stress. Heat stress experiments designed to repeat the observations made by

Eguchi et al. (1996) initially gave conflicting and poorly reproducible results. In order to examine the effect in detail a range of heat stress temperature on the survival of S. alaskensis, exponentially growing batch cultures were exposed to heat stress in the form of a shift from the growth temperature (30˚C) to 52˚C, 54˚C, 55.2˚C, 56˚C and 58˚C and continuous exposure at those temperatures for up to 1 h. To minimise variation the heat stress experiments were carried out simultaneously in thin-walled

PCR tubes in the heating block of thermal cycler with a temperature gradient

98 Chapter 4

function. The exact temperature was monitored using a calibrated temperature probe and a mercury thermometer.

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01 % Survival

1.0E-02

1.0E-03

1.0E-04 0 2000 4000 6000 8000 10000

Ultraviolet dose (Jm-2)

FIGURE 4.7. Survival of batch grown S. alaskensis cells following exposure to UV radiation. UV doses ranged from 0 to 10,000 J m-2. Survival was monitored by plating serial dilutions on ASWG ()andMNB().

99 Chapter 4

No loss of viability was recorded when S. alaskensis cultures were exposed to temperatures up to 54˚C (Fig 4.9). However, after continual exposure to 58˚C for 1 hr the survival fraction dropped to below 0.0001%. Survival after exposure to 55.2˚C and 56˚C for 1 h was 5.0% and 0.005% respectively. These results demonstrate that, over a very narrow temperature range, heat stress can have profound effect on the survival of S.alaskensis . This observation may account for the previously encountered variation in results obtained by different researchers and on different occasions and those published by Eguchi et al. (1996).

When the effect of glucose limitation on heat stress was determined with the methods defined above at a temperature of 56˚C for up to 60 min. No clear trends were evident after cultures grown at different rates in a glucose limited chemostat were subjected to heat stress at 56˚C (results not shown). No clear trends were evident and there were some discrepancies between the CFU ml-1 values obtained on

MNB and ASWG solid medium (results not shown).

S. alaskensis cells grown under any condition were remarkably resistant to freeze- thaw stress and cold stress at 4˚C. Irrespective of the preceding growth conditions loss of viability after freeze-thaw was not significant. More than 90% of cells subjected to each stress remained viable (results not shown).

100 Chapter 4

1.0E+02

1.0E+01

1.0E+00

1.0E-01

% Survival 1.0E-02

1.0E-03

1.0E-04 0 0.05 0.1 0.15 0.2 0.25

Dilution rate (h-1)

FIGURE 4.8. Percent survival of batch and chemostat grown S. alaskensis following exposure to 4000 Jm-2 ultraviolet radiation. Samples were taken directly from steady state glucose-limited chemostats at =0.03, 0.06, 0.08, 0.16 and 0.17 h-1 and from batch culture at max (0.21 h-1). Experiments were performed twice and colony forming units counted using the drop-plate method on MNB () and ASWG (). A standard deviation of less than 20% was observed for each time point.

101 Chapter 4

1.0E+03

1.0E+02

1.0E+01

1.0E+00

1.0E-01

1.0E-02

1.0E-03 Percent Survival 1.0E-04

1.0E-05

1.0E-06 50 52 54 56 58 60 Temperature (˚C)

FIGURE 4.9. Percent survival of exponentially growing S. alaskensis following heat stress. Samples were taken directly from log-phase batch culture exposure to 52˚C, 54˚C, 55.2˚C, 56˚C and 58˚C for 1 h. Experiments were performed at least three times and colony forming units counted using the drop-plate method on MNB and ASWG.

102 Chapter 4

4.3.9 The possible role of cellular pigments in stress resistance of S. alaskensis

Distinct differences were observed in the colour of S. alaskensis cultures when growing at different rates in nutrient-limited chemostats. When observed at equivalent optical density and cell numbers, slowly growing glucose- and amino acid-limited cultures appeared a darker hue of yellow when compared with batch and chemostat cultures growing at faster rates. Other factors, such as growth temperature and growth substrates also qualitatively affected the colour of cultures. Since the yellow pigmentation of other Sphingomonodaceae had been identified as the caroteoind nostoxanthin (Jenkins et al., 1979, Davison et al., 1999), and carotenoids are believed to play a major role in the protection of proteins and DNA against cellular damage resulting from UV radiation and free radicals in some organisms

(Jenkins et al., 1979, Ourisson and Natatani, 1990, Fong et al., 2001) we rationalised that the abundance of the cellular pigment at different growth rates and conditions may be linked to the observed patterns of stress resistance in S. alaskensis.

Ultraviolet/visible absorption spectra of cellular methanol extracts exhibited three peaked maxima characteristic of carotenoids (Fig. 4.10a). The observed characteristic absorption maxima at 420, 447 and 477 nm and direct spectrophotometric comparison with the pigment extracted from Sphingomonas paucimobilis S88 identify the major pigment component of S. alaskensis as nostoxanthin (Fig 4.10, Fig 4.11). Characteristic cis maxima that can be of aid in identifying carotenoid cis isomers by spectrophotometry were not visible in crude

103 Chapter 4

cellular extracts due to the presence of a compound with large absorbance at 230 nm

(Fig 4.10).

4.3.10 The effect of growth conditions and nutrient limitation on the abundance

of nostoxanthin of S. alaskensis

To quantitatively determine whether carotenoid levels correlated with cellular traits linked to growth conditions (eg. hydrogen peroxide resistance), the abundance of carotenoids was measured in cellular methanol extracts of S. alaskensis grown at various rates of growth in glucose limited chemostats and in batch cultures (Fig.

4.12). The relative abundance of carotenoids varied by about three-fold with glucose- limited cultures grown at low rates of growth producing the greatest amount, cultures grown at high dilution rates produced intermediate levels while batch cultures produced the least. During glucose limited growth the gradual change in the overall abundance of carotenoids observed throughout the range of growth rates tested and the less than two-fold difference in carotenoid levels between cells grown at high and low rates of growth did not correlate with the 10,000-fold difference in hydrogen peroxide resistance. These data indicate that the overall abundance of carotenoids is not responsible for the bimodal response to hydrogen peroxide observed in S. alaskensis, however, they do not exclude the role of carotenoids from the intrinsically high levels of resistance to oxidative stress and ultraviolet radiation.

104 Chapter 4

2.0

A A

1.5

1.0

Absorbance (A)

0.5

0.0 200 300 400 500

0.6 B 0.5

0.4

0.3

Absorbance (A) 0.2

0.1

0.0 350 400 450 500 550

Wavelength (nm)

FIGURE 4.10 Absorbance spectra of cellular methanol extracts of S. alaskensis.(A) from 190 nm to 600 nm and (B) from 350 nm to 550 nm.

FIGURE 4.11 Molecular structure of nostoxanthin.

105 Chapter 4

0.14

0.12

0.1

0.08

0.06 Absorbance (A)

0.04

0.02 0 0.05 0.1 0.15 0.2

Dilution Rate (h-1)

FIGURE 4.12 Relative abundance of carotenoids extracted from S. alaskensis grown in glucose-limited chemostats and batch cultures. Cultures were grown at various specific rates of growth in glucose limited chemostats (closed symbols) and at max batch culture (open symbols). Cultures were grown at ambient light levels at 30˚C. Total carotenoids were extracted from 3.0 x 109 cells and relative abundance was measured by spectrophotometry at 447 nm () and 477 nm (). Standard error bars represent the standard error of extracts taken from replicate cultures.

106 Chapter 4

4.4 DISCUSSION

4.4.1 Growth rate control of hydrogen peroxide resistance

We have previously found that cultures of S. alaskensis grown at 25 ˚C in a glucose- limited chemostat ( = 0.03 h-1) were more resistant to hydrogen peroxide than cultures grown in batch ( = 0.16 h-1) (Eguchi et al., 1996). In the present study we examined whether a link existed between the actual specific growth rate of the culture and hydrogen peroxide resistance, and whether the resistance of the cells was significantly affected by the mode of growth, e.g.: carbon or nitrogen limited growth in chemostats and growth under conditions of nutrient excess in batch. It is indeed apparent that hydrogen peroxide resistance in S. alaskensis is strongly influenced by the specific growth rate of the cultures when glucose or mixed amino acids are the limiting substrates (Fig. 4.1 and Fig 4.4b), however growth-rate control of hydrogen peroxide-stress resistance was not apparent under nitrogen limitation (Fig. 4.4b).

Thus, comparison of stress-survival in cells grown under carbon/energy limitation with ammonium limited grown cells at comparable dilution rates in chemostats (i.e.: approx. 0.03 – 0.05 h-1 and0.14–0.16h-1), clearly shows that the specific rate of growth is not the only major determinant for the level of stress resistance in S. alaskensis. These results indicate that hydrogen peroxide stress resistance in S. alaskensis is influenced directly or indirectly by both the nature of the limiting substrate (carbon/energy vs nitrogen) and by the extent of nutrient-limitation imposed in carbon and/or energy limited chemostat cultures. In contrast, exponentially growing or starved cells from batch growth had equivalent levels of resistance to carbon-limited, chemostat grown cells at high dilution rates.

107 Chapter 4

In order to establish whether the physiological responses were unique for S. alaskensis, we also examined the hydrogen peroxide resistance of V. angustum S14 in response to specific growth rate in glucose-limited chemostats. The rate of growth had some influence on the peroxide-stress survival, however starved and exponentially growing cultures represented the extremes of peroxide-stress resistance and sensitivity, respectively. These observations are in line with several earlier observations which all indicate that starvation evokes a strong cross-protection against heat or peroxide stress in V. angustum S14 and other phylogenetically unrelated organisms (Golovlev, 1999, Givskov et al., 1994, Jenkins et al., 1988,

Jouper-Jaan et al., 1992, Nyström et al., 1992, Preyer and Oliver, 1993). In support of this, E. coli grown in batch and glucose-limited chemostats had a similar response to hydrogen peroxide stress as was observed for V. angustum S14. This emphasizes the uniqueness of the response observed in S. alaskensis in which starvation does not elicit any cross-protection to hydrogen peroxide or heat stress (Eguchi et al., 1996), whereas substrate-limited growth rates of less than approximately 75% of max result in highly increased hydrogen peroxide resistance.

A striking characteristic of the response in S. alaskensis was that the increased stress- survival was not gradual, but that lowering the growth rate below 0.13 h-1 seemed to act like a switch. This contrasts the general picture which emerges from most of the earlier observations on the influence of changes in specific growth rates of cells in carbon/energy limited chemostats: all of which point to gradual changes in enzyme levels (Matin, 1981, Harder and Dijkhuizen, 1983, Tempest and Neijssel, 1978,

Notley and Ferenci, 1996), viability (Tempest et al., 1967, Moyer and Morita, 1989a,

108 Chapter 4

Gottschal, 1990, Poolman et al., 1987), cell sizes (Maalöe and Kjeldgaard, 1966,

Matin and Veldkamp, 1978, Koch, 1979, Moyer and Morita, 1989a), concentrations of soluble and structural cell components (Maalöe and Kjeldgaard, 1966, Harder and

Dijkhuizen, 1983, Liu and Ferenci, 1998, Kemp et al., 1993, Holms, 1996, Tweedale et al., 1998) and the abundance and influence of stationary phase sigma factors

(Notley and Ferenci, 1996, Schweder et al., 1999).

4.4.2 Possible mechanisms of growth rate control of hydrogen peroxide

resistance.

The phenotypic responses in S. alaskensis are consistent with a regulatory cascade where very small changes in the concentration of effector molecules and/or proteins become amplified through the cascade. Such regulatory events may be mediated by global regulators such as rpoS, oxyR and soxRS and other mechanisms such as DNA methylation and stringent control, which are important in oxidative stress responses in E. coli (Farr and Kogoma, 1991, Storz and Imlay, 1999).

The fact that the growth-rate dependent, hydrogen-peroxide resistance is linked to carbon and/or energy-limitation (Fig. 4.1 and Fig 4.4b) but not nitrogen-limitation

(Fig. 4.4a) indicates that the sensing mechanism involves a response to the flux through a metabolic pathway related to carbon uptake or carbon/energy metabolism.

When uptake of glucose is restricted in a glucose-limited chemostat, acetate excretion in E. coli reduces until it becomes zero at a growth rate of 0.72 h-1 (Holms,

1996). While a reduction in growth rate leads to a gradual change in flux through central metabolic pathways, there comes a point when acetate is no longer produced.

109 Chapter 4

It is possible that a similar change in flux through carbon metabolism in S. alaskensis provides the signal that leads to the sudden change in stress resistance.

The mechanism that leads to dramatically enhanced hydrogen peroxide resistance in

S. alaskensis has not been determined. However, catalase activity does not appear to be linked to the resistance state of the cells. Total catalase activity varied by about 6- fold (Table 4.1) and did not correlate with stress resistance. In their review on the regulation of enzyme synthesis in bacteria grown in chemostats, Matin (1981) reported five types of changes in enzyme activity in response to growth rate.

Interestingly, while about 50% of enzymes exhibited increased activity with decreasing growth rate (Matin, 1981), in E. coli, superoxide dismutase activity increased with growth rate, peroxidase activity decreased with growth rate and catalase activity was invariant until the growth rate exceeded 0.4 h-1, when it then declined (Hassan and Fridovich, 1977).

While these studies in E. coli and our studies on S. alaskensis indicate that regulation of catalase expression appears to be largely independent of growth rate, the involvement of the katC locus in E. coli illustrates the complexity by which catalase gene expression may be regulated. Volkert et al. (1994) have reported that the katC locus is responsible for the sensitivity of wild-type strains to hydrogen peroxide.

Deletion of this locus in an argF-lacZ strain, or interruption of the locus with Tn9, apparently leads to a dramatic increase in hydrogen peroxide resistance: ~1,000-fold greater survival after 25 min exposure to 150 mM hydrogen peroxide compared to wild-type. Interestingly, while the increased resistance is starvation dependent, and requires functional katE and katF genes, catalase activity and katE expression is not

110 Chapter 4

elevated in katC mutant strains compared to wild-type. The authors speculate that the product of the katC gene (IS1B-IS30B fusion), may affect a function of KatF that involves an activity other than catalase. While the function of katC is unknown, it provides the precedent for a genetic element that affects the hydrogen peroxide resistance of isogenic strains by at least 3-orders of magnitude.

Hydrogen peroxide and other reactive oxygen species are capable of causing damage to DNA, RNA, protein and lipids (reviewed in Storz and Imlay, 1999). Catalase activity is only one means of reducing the damaging effects of hydrogen peroxide.

Factors that may be involved in the increased resistance of slowly growing cells may include increased levels of detoxifying enzymes such as proteins that reduce disulfide bridges caused by oxidative stress (e.g. glutathione reductase), organic hydroperoxidases other than catalase (e.g. Ahp) or nucleic-acid binding proteins (e.g.

Dps). In addition, the cell wall or cytoplasmic membrane may be modified to reduce hydrogen peroxide penetration into the cell (Subczynski et al., 1991).

The slow degradation rate of hydrogen peroxide, as seen by the slow disappearance of hydrogen peroxide in catalase assays, may provide an important clue. A slow conversion rate of hydrogen peroxide into more reactive intermediates, such as the superoxide or hydroperoxyl radicals, may explain how a small change in catalase expression can account for the marked difference in survival. Since the production of radicals through the Fenton reaction is dependent on free Fe2+ ions in the medium, periplasm or cytoplasm, an altered availability of Fe2+ at low rates of growth could be a determining factor that could be explored further. A reduced free Fe2+ availability may be a result of the production of Fe sequestering proteins such as

111 Chapter 4

ferritin, siderophores, enhanced uptake though a high affinity system and an increased demand for Fe2+ as a cofactor in enzymes.

The superoxide radical sources paraquat and menadione were also tested. Chemostat grown cultures were more than 85% resistant to 500 mM paraquat for 15 min (A.

Goodchild Hons. Thesis, UNSW, 2001). The results did not mirror the effect of hydrogen peroxide but are worth further investigation.

4.4.3 Resistance to ultraviolet radiation

It has previously been shown that S. alaskensis is highly resistant to UV-B radiation and t does not accumulate cyclo-pyrimidine dimers (CPDs) after different UV-B doses (Joux et al., 1999). The authors found that the total amount of CPDs in S. alaskensis was four-fold less than the amounts in the other marine bacteria at a UVB dose of 2000 J m-2 and suggest that this may be due to a constitutive photo-protective mechanism. While UV-B causes DNA-strand breakage, the protective mechanism may also repair damage caused by reactive oxygen species. In the present study we examined the effect of growth rate on the resistance of S. alaskensis to broad spectrum UV radiation.

The resistance of S. alaskensis to UVR was relatively high regardless of growth conditions (Figure 4.8). In contrast to the bimodal response to hydrogen peroxide, a very weak trend of increasing UVR resistance with increasing growth rate was discerned. The different response to UVR suggest that the factor(s) responsible for increased resistance to hydrogen peroxide at low dilution rates do not provide cross

112 Chapter 4

protection against ultraviolet radiation. The levels of resistance measured were equivalent to those reported by Joux et al. (1999), indicating that the UV-A component of the UVR dose used in this study has little effect, either positive or negative, on the survival of this organism. The ability to withstand high levels of

UV-B may be an important factor that allows S. alaskensis to grow and divide throughout daylight hours in the surface waters of the ocean. Possible mechanisms of protection against UV-B include nucleic-acid binding proteins (e.g. Dps) and the production of UV-B screening compounds (Ferguson et al., 1998).

4.4.4 Starvation vs low growth rate induction of peroxide stress resistance

In a recent review Nyström evokes the free radical hypothesis of aging to explain the progressive deterioration (death) of starved bacterial cells (Nyström, 1999). That is, loss of viability is the eventual result of the progressive accumulation of oxidative damage to cellular components. Increases in the levels of oxidative stress defence proteins have been well documented for starved bacterial cells (reviewed in Nyström,

1999). Recent data support the idea that the levels of protective proteins are upregulated to specifically target endogenously derived oxidative stresses caused by ongoing respiration in starved cells rather than to protect against a future oxidative burst associated with the outgrowth response (Nyström et al., 1996). Since starved or slowly growing (i.e. energy/nutrient limited) cells possess a diminished capacity to replace or dilute damaged cellular components by new synthesis it seems reasonable for S. alaskensis to make considerable investment in protective and repair mechanisms.

113 Chapter 4

In almost all cases starvation survival, and in some cases starvation-induced cross protection, is dependent upon the substrate for which the culture is starved. V. angustum S14 starvation-induced survival, miniaturisation and stress resistance is specific for carbon, but not nitrogen or phosphorus starvation. E. coli viability is more sensitive to phosphate starvation than to carbon or nitrogen (Davis et al., 1986) while porin regulation in response to glucose-limitation is not observed in nitrogen limited cultures (Liu and Ferenci, 1998). An explanation based on the ecology of marine bacteria may be related to the fact that most pelagic heterotrophic bacteria are limited by the availability of carbon (Kirchman, 1990, Church et al., 2000) and adaptive responses are geared for carbon starvation.

4.4.5 Possible role of cell membrane and pigments in overall stress resistance

Carotenoids are a structurally diverse group of lipophilic isoprenoids consisting of a

C-40 methyl branched hydrocarbon backbone. They often carry cyclised and derivatised hydrophilic terminal structures (Britton, 1995). Due to their overall lipophilic nature carotenoids are most often found embedded in membranes. The later stages of carotenoid synthesis occur on or within the membrane while shorter precursors, such as phytoene, may be found in the cytoplasm. Carotenoids account for the yellow pigmentation of S. alaskensis and may account for a significant component of the membranes. Spectrophotometry indicates that the major pigment component of S. alaskensis is the carotenoid nostoxanthin ((2R, 3R, 2’R, 3’R)-,- carotene-2,3,2’,3’-tetrol). Nostoxanthin is notable for its role as a photoprotectant and accessory pigment in the photosystem antenna complexes of some cyanobacteria and for its occurrence in photosynthetic and non photosynthetic members of the

Sphingomonas and Erythrobacter genera. 114 Chapter 4

Carotenoids have been shown to be effective scavengers of reactive oxygen species and to protect cellular components, including DNA, from damage (Jenkins et al,

1979, Ourisson and Natatani, 1990, Bridges and Timms, 1998). Nostoxanthin is structurally very similar to other naturally occurring carotenoids such as zeaxanthin and astaxanthin which are efficient antioxidants and known to inhibit lipid peroxidation significantly (Woodall et al., 1997, Goto et al., 2001). The efficient antioxidant activity of astaxanthin is thought to be due in part to the unique terminal ring structureswhich are responsible for the initial quenching of radicals at the membrane surface, while the conjugated polyene moiety is capable of quenching radicals within the membrane after surface quenching has been overwhelmed.

Carotenoid levels in S. alaskensis increase about two-fold with decreasing growth rate in glucose limited chemostats (Fig 4.12), however, the carotenoid levels do not correlate with the observed bimodal response to hydrogen peroxide. These data clearly indicate that the carotenoid abundance per se is not responsible for the increased level of resistance to hydrogen peroxide below a specific growth rate of

0.133 h-1. The methods used in this study are not suitable for detecting isomers or subtle chemical modifications of nostoxanthin. Goto et al. (2001), and Sandmann et al. (1998), suggest that chemical modifications, such as glycosylation, and the cellular localisation of carotenoids may have an important influence on the effectiveness and nature of their protective activities.

Polar carotenoids, such as zeaxanthin, may play a role in protecting the cell from

UV-B radiation (Hoshino et al., 1993, Sandmann et al., 1998, Götz et al., 1999). In S.

115 Chapter 4

alaskensis there is a weak negative correlation of the overall abundance of carotenoids with the resistance to UVR, suggesting that the cellular pigment does not offer protection by direct screening of UV radiation.

Experimental data also suggest that carotenoids act as membrane reinforcers, playing a role in membrane strength and stability (Ourisson and Natatani, 1990, Goto et al.,

2001). Since the length of the hydrophobic backbone corresponds to the width of a membrane and the cyclised and derivatised ends are hydrophilic in nature carotenoids bearing two highly polar distally placed groupings can be easily incorporated into phospholipid membranes and may be regarded as membrane rivets.

Due to the small size and large surface area to volume ratio and the requirements for nutrient scavenging and electron transport machinery, the membranes of S. alaskensis are likely to be rich in protein, carotenoids may play an important structural role in ensuring membrane integrity.

The presence of carotenoids may preclude the entry of reactive oxygen species and precursors, such as hydrogen peroxide, into the cell or limit the diffusion of reactive species throughout the cell (Subczynski et al., 1991, Britton, 1995). If carotenoids can change the effectiveness of the membrane as a barrier to water and oxygen it seems reasonable to expect that the diffusion of larger, more polar molecules, such as hydrogen peroxide, would be similarly decreased. It is also important to consider that the profound reactivity of free-radicals, such as the hydroxyl radical (HO-), limits the effective distance that they may travel from their source before reacting with a biological molecule. In this regard, localisation of carotenoids in the cell membrane

116 Chapter 4

is optimal for effective scavenging of reactive oxygen species generated by the respiratory electron transport chain.

The possible roles for nostoxanthinin the intrinsic stress resisitance of S. alaskensis warrants further investigation, possibly through the use of chemical caotenoid biosynthesis inhibitors (e.g. nicotine) or by employing random or directed mutagenesis techniques.

117 Chapter 5

5AN ASSESSMENT OF TWO DIMENSIONAL

PROTEIN REFERENCE MAPS FROM S.

ALASKENSIS RB2256

118 Chapter 5

5.1 BACKGROUND

A range of methods have been developed and are widely used for comprehensive analyses of gene expression in complex biological systems. Generally these methods assess global gene expression at the level of either or . The collective technologies used for the characterisation of the latter are referred to as proteomics, and the former as transcriptomics. The ‘proteome’ consists of all proteins expressed by an organism under a given set of conditions and therefore represents the functional complement of the genome. Proteome analysis is commonly achieved by a combination of sample fractionation and two-dimensional electrophoresis (2DE) for separation and visualisation of mixtures of proteins and mass spectrometry (MS) for protein identification.

In 2DE, proteins are first separated on the basis of charge using iso-electric focussing

(IEF). IEF is achieved with the use of commercially available immobilised pH gradients (IPGs) that are available for a variety of pH ranges. Broad range IPGs, e.g. pH 3-10 can provide a broad overview of expressed proteins while narrow range

IPGs (pH 4-7 and 5.0 –6.0) can be used to increase the number of resolved protein spots within a given interval by improving the resolution of co-migrating spots and accommodating larger sample loads, enabling the detection of less abundant proteins.

Following IEF, proteins are then separated on the basis of their molecular weight by sodium dodecyl sulfate (SDS) poly-acrylamide gel electrophoresis. The resulting

2DE profiles can then be visualised by a number of methods: two commonly used methods are silver-staining and autoradiography of radioactively labelled samples.

The characteristics of each of these methods are fundamentally different, for

119 Chapter 5

example, silver-staining is used to visualise all proteins present in a mixture, while

‘pulse’-labelling with radioactively labelled amino acids can be used to detect newly synthesised proteins in response to controlled environmental changes.

By examining protein profiles generated by 2DE, gene expression can be recorded and compared for any biological condition that can be experimentally monitored, for example, throughout the growth phases of an organism, in stressed vs non-stressed states and in during nutrient limited growth, vs unlimited growth. By comparing the protein expression patterns under a variety of different conditions those proteins that are co-regulated can be classified into groups or ‘regulons’ which represent sets of genes that function in the same process (Petersohn et al., 2001, Giard et al., 2001).

By this approach the complex and overlapping expression patterns of protein

‘regulons’ can be defined, and protein responders can be identified for further analysis. Alternatively, those protein spots that are up- or down- regulated by a single set of conditions can be identified and targeted for further characterisation.

There are numerous examples in the literature where proteomics has been used to identify protein spots associated with certain cellular states. For example, Oosthuizen et al., (2002) observed 24 out of 345 (or 7%) protein spots unique to one of two periods during biofilm formation by Bacillus cereus cells, Grünenfelder et al. (2001) observed 144 out 979 protein spots with altered levels of expression throughout the cell cycle of C. crescentus and Östling et al. (1996) observed 301 out of 1,700 protein spots with altered expression levels in response to starvation of Vibrio angustum S14.

120 Chapter 5

In addition to defining the expression patterns of particular protein spots, 2DE gels are also valuable for isolating specific proteins for identification. Advances in mass spectrometry and the rapid expansion of sequence databases have dramatically enhanced the ability to characterise proteins taken directly from 2DE gels. Where comprehensive genome sequence data is not available for a particular organism, proteins taken from 2DE gels can be identified by comparing analytical data to database sequences from other species (Wilkins et al., 1998). Useful protein parameters that are conserved across species boundaries include primary sequence, amino acid composition, peptide masses and molecular weight (Courchesne and

Patterson, 1997, Wilkins et al., 1998, Cordwell et al., 1997).

The results of proteomics projects serve to illustrate the usefulness of a global approach, and how it can be combined with other methodologies in order to achieve an improved understanding of the means by which bacteria adapt to alterations in environmental conditions. The overall strategy used in this study is particularly useful for studies of bacteria for which efficient genetic tools, background genotypes or genome sequence are unavailable. This chapter describes the protocols developed to generate 2DE protein profiles for S. alaskensis growing in chemostat cultures.

Two different methods of visualisation are evaluated for generating quantitative analytical images: radioactive pulse labelling with [35S] L-methionine/cysteine and silver staining. The methods developed and the 2DE profiles generated provide the basis for the examination of gene expression in response to nutrient limitation in chemostat cultures (Chapter 6).

121 Chapter 5

5.2 MATERIALS AND METHODS

5.2.1 Biomass sampling and preservation

For cultivation in chemostat cultures S. alaskensis RB2256 was grown in ASW with 9 -1 3.0 mM glucose (OD433 0.9, 3 x 10 cells ml ). Aliquots (50 ml) for analysis of total protein were collected from the chemostat and cells were collected by centrifugation (8,000 g, 4˚C). Excess salt was removed from the cell pellet by resuspension in ice- cold 0.2 M sucrose and the cells were collected by centrifugation (8,000 g, 4˚C). After the supernatant was discarded samples were either processed immediately or the pellet resuspended in 100 l of 0.2 M sucrose, transferred to 8 cm glass vials and freeze-dried with an Edwards Micro Modulyo for long term storage in the dark at room temperature.

5.2.2 Pulse labelling of chemostat cultures

Glucose limited cultures of S. alaskensis growing in steady state 50 ml chemostats 9 -1 35 (OD433 0.3, 10 cells ml ) were labelled with [ S] L-methionine/cysteine (10.0 l, 1,175 Ci mmol-1,10mCiml-1, ICN Pharmaceuticals, Aurora, OH, USA). The labelling time was 1 h at 30˚C and the incorporation of radiolabel was monitored throughout. After 1 h, unincorporated label was diluted with 500 l of 0.1 M ice-cold L-methionine and the entire culture was harvested from the chemostat. Cells were collected by centrifugation at 8,000 g for 8 min, washed in and equal volume of ice-cold 0.2 M sucrose and recollected by centrifugation twice. Finally, any remaining salt was removed from the pellet by washing with 2 ml ice-cold methanol. Samples were either processed immediately or preserved for up to two months at – 80˚C.

5.2.3 Sample preparation

Cell pellets were resuspended in 200 l rehydration buffer (8 M urea, 0.1 M DTT, 40 mM TrisCl pH 8.8, 1.2% v/v Pharmolytes pH 3-7, 4% w/v CHAPS). Non radioactive samples were sonicated on ice for 2-5 min with a Branson Sonifier on a 30% duty 122 Chapter 5

cycle and a setting of 3.5. radioactive samples were sonicated by floating a microfuge tube containing the sample in ice-cold water in a NEY ultrasonik 300, at medium setting for 5 min. To remove nucleic acids, 8 l of nuclease buffer (1 mg of DNAse I ml-1,0.25mgofRnaseml-1, 24 mM Tris base, 476 mM Tris HCl, 50 mM

MgCl2) was added to each sample and incubated on ice for 20 min. Samples were centrifuged at 14,000 rpm for 30 min at 4oC to remove cell debris and the supernatant retained for electrophoresis. For both methods the absence of a substantial turbid pellet after centrifugation indicated that no unbroken cells remained and therefore that cellular lysis was satisfactory.

To determine radioactivity levels (disintegrations min-1, dpm) duplicate aliquots (10 l) of each labelled sample was added to 10 ml of scintillation cocktail (Wallac Hi Safe 3) and dpm measured with a Tri-carb 2100 Pakard liquid scintillation analyser. Protein concentration was determined using the Bicinchoninic Acid (BCA) kit (SIGMA) with BSA as a standard.

5.2.4 Isoelectric focusing

An aliquot containing 75 g of protein or 5 x 106 dpm was added to rehydration buffer in a total volume of 500 l and used to rehydrate 18 cm Immobiline DryStrips, pH 4-7 (Pharmacia) for a minimum of 6 h. Iso-electric focusing was performed using a Multiphor II system (Pharmacia) at 15°C programmed for 0.5 h sequential intervals of 300 V, 1 kV and 2.5 kV, followed by 17 h at 3.5 kV, for a total of 61,400 Vh.

5.2.5 Equilibration and SDS-PAGE

IPGs were equilibrated by washing in fresh equilibration buffer (6 M urea, 2% w/v SDS, 20% glycerol, 0.375 M Tris-base, pH 8.8) plus 2% DTT (Herbert et al., 1997) for 20 min with gentle shaking. After the wash, the solution was discarded and replaced for 10 min with equilibration buffer supplemented with 2.5% (w/v) acrylamide. SDS-PAGE was performed in 11.5% (v/v) polyacrylamide (0.8% bis) gels using a Millipore Investigator system. Gels were electrophoresed at 1,600 mW 123 Chapter 5

per gel at 15oC for five and a half hours or until the bromophenol blue front reached the bottom of the gels. Alternatively, gels were electrophoresed using a PROTEAN xi (BioRad) at 20 mA per gel at 18˚C until the bromophenol blue front reached the bottom of the gel. Following the completion of the run labelled gels were fixed in a solution of 40 % (v/v) ethanol and 10 % (v/v) acetic acid. Non-labelled gels were fixed in a solution containing 40% (v/v) methanol and 10% (v/v) acetic acid.

5.2.6 Gel staining

Analytical gels were stained with a sensitive ammoniacal-silver method (Rabilloud et al., 1994). Fixed gels were transferred to a second fixer (4.5% w/v sodium acetate, 30% v/v methanol, 0.5% v/v glutaraldehyde) for 1 h followed by 3 x 15 min washes in 400 ml MilliQ. Washed gels were incubated in napthalene-disulfonic acid (NDS) solution (0.05% w/v, 400 ml) for at least 1 h, washed in fresh MilliQ (4 x 20 min, 400 ml) and stained for 1 h (0.6% (w/v) silver nitrate, 1.5% (v/v) ammonia and 0.08% (w/v) sodium hydroxide, 200 ml). Silver-stained gels were rinsed in MilliQ (2 x 10 min, 400 ml) prior to development (5-10 min, 0.01% w/v citric acid, 0.1% v/v formaldehyde). Development was stopped with 5% acetic acid.

5.2.7 Image acquisition and spot detection

Digital images of silver-stained gels were obtained with a model GS-700 Imaging Densitometer (Biorad). Radioactive gels were dried at 80˚C for 1 h in a model 583 gel dryer (Biorad) and then exposed to a CS phosphor screen (Biorad) for 96 h. Gel images were extracted from CS phosphor screens with a model GS 525 Molecular Imager (Biorad). Image files were obtained with the aid of Multi-Analyst software (BioRad) and were exported as 16 bit tiff. Digitised images were analysed by the Z3 software package (Compugen). Imported 16 bit tiff images were decimated (factor 2) and noise reduction was applied (radius 3). The critical thresholds for spot detection were as follows: minimum fill ratio 0.33, spot detector dynamic range = 5 for images of silver stained gels and 7 for images of radiolabelled gels, minimum spot area = 20, minimum spot contrast = 25 min, confidence limit = 0.95 (student t-test). Spot matching utilised semi-automatic registration (alignment), maximum fit calibration

124 Chapter 5

and a confidence limit of 0.95 (student t-test). Replicate gels from the same condition were registered (aligned) and compiled into a single or multiple Raw Master Gel (RMG) images for comparative analysis. Spot matchings, revealed from RMG analyses, were manually edited, spot by spot, gel by gel, for differentially expressed proteins.

2.3 Image analysis

For comparative image analysis, statistical data were acquired using Z3 software (Compugen, Israel). The computational model used in Z3 software is described in a recent paper depicting Compugen’s large-scale analysis software, Z4000 (Rubinfield et al., 2003). Imported 16 bit tiff images were decimated (factor 2) and noise reduction was applied (radius 3). The critical thresholds for spot detection were as follows: minimum fill ratio 0.33, spot detector dynamic range, 5 for images of silver stained gels and 7 for images of radiolabelled gels; minimum spot area, 20, minimum spot contrast, 25 min, confidence limit, 0.95 (student t-test). Spot matching utilised semi-automatic registration (alignment), maximum fit calibration and a confidence limit of 0.95 (student t-test). Replicate gels from the same condition were registered (aligned) and compiled into a single raw master gel (RMG) image for comparative analysis. All RMGs were generated from gels produced from 2 independent chemostat cultures for each growth rate. For analysis of radiolabelled cultures, RMG images were generated from a total of 5 and 4 replicate gels for low and high growth rates, respectively. For unlabelled cultures, replicates were generated from 4 and 5 replicate gels of low and high growth rates, respectively. To determine differential spot intensity, matched spots from RMG analyses were manually edited spot-by-spot and gel-by-gel. Spot intensity was expressed as ppm which is the ratio between the intensity of a single spot and the total intensity of all spots on a gel, which takes into account variability due to staining and detection. Maximum fit calibration was used to maximise the number of spots where the differential spot intensity is as close to 1 as possible. This method of calibration corrects for overall image differences caused by variations in the electrophoresis and staining conditions and thereby serves to identify those proteins with the most significant differences in expression levels. Statistical analyses (student t-test, 95% confidence interval) were performed on at

125 Chapter 5

least 3 gels from each condition to determine spots showing reproducible n-fold changes.

126 Chapter 5

5.3 RESULTS AND DISCUSSION

5.3.1 Characteristics of silver stained and radiolabelled 2DE images

Two-dimensional gels were prepared for glucose limited chemostat cultures of S. alaskensis. Protein spots were detected on digitised images with the aid of Z3 software (Compugen, Israel). The quality of individual 2DE gel images can be assessed on the basis of the number individual spots that are resolved. More than

1,800 individual protein spots could be visualised on a single silver stained gel

(Table 5.1). The maximum resolution was obtained on a silver stained 2DE gel with an isoelectric focussing range of pH 4-7 and a protein load of 75 g. Higher protein loads decreased the resolution of gels in both dimensions due to the co-migration of spots. It is notable that maximum resolution was obtained when contaminating salts were removed from the culture sample by 0.2 M sucrose washes and further removed from the crude extract by methanol wash. No differences in protein profiles were observed between gels produced from freshly prepared or freeze-dried protein samples, demonstrating that freeze drying was an effective method for preparing samples for transport between laboratories and long term storage.

A total of four electrophoretic gels were prepared from total protein taken from three independent chemostat cultures growing at = 0.03 h-1 (Fig 5.1). The total number of spots matched in at least three out of four silver stained gel images was 1,020. The number of matched spots is low in comparison to the total number of spots that can be visualised. This difference is due to spots with low intensities that were not consistently detected and spots at the extremities of the gels that were not consistently resolved. A raw master image was compiled by merging the four 127 Chapter 5

electrophoretic gel images into a single image and the 1,020 reproducible spots were analysed further. Protein spots were distributed evenly throughout the pH (4.0 –7.0) and Mr (10–150 kDa) range. Spot intensity was measured as parts per million (ppm) relative to the total intensity of all detected spots across the entire image. Spot intensity varied from 220–9881 ppm. This is reflected by the fact that the ten most abundant proteins accounted for more than 14.7% of the total intensity measured across the entire image. The top 100 most abundant proteins account for approximately 47.5% of the total protein detected and quantified on a gel with a pI interval 4-7.

35 Radiolabelled proteins were prepared by directly injecting a labelling mix of [ S] L- methionine into each chemostat culture during steady-state. The relative incorporation of radiolabel was monitored throughout a 60 min labelling period at high, low and medium rates of growth (Figure 5.2). Most (> 60 %) of the radioactivity was incorporated into acid insoluble material, indicating uptake into the cell and incorporation into macromolecules, within the first five minutes regardless of the culture conditions. After 60 min of labelling at least 2.4 x 106 dpm was incorporated into cellular material for each culture. Labelling efficiency was highest for cultures grown at low rates of growth, which averaged 50% higher dpm values, in comparison with medium and high growth rate cultures at each time point monitored.

128 Chapter 5

35 Figure 5.1. The incorporation of [ S] L-methionine for cultures grown at different glucose limited rates in chemostat cultures. Labelled methionine was injected directly into chemostat cultures growing at 0.03 (), 0.13 () and 0.18 h-1 (). Duplicate samples were taken at 0, 5, 10 and 60 m time intervals and incorporation into acid-insoluble material was measured by scintillation.

35 Five electrophoretic gels were prepared from [ S] L-methionine pulse labelled cultures growing at = 0.03 h-1 (Fig. 5.3). In general the number of protein spots detected on radiolabelled gel images of was slightly less than the number detected on silver stained gels (~1,600 vs ~1,800). The total number of matched protein spots in at least three out of five gels was 870, which is slightly lower in comparison to the number on silver stained gels (Table 5.1). This was due largely to the number of spots that did not resolve consistently at the extremities of the gels and to the

129 Chapter 5

decrease in the total number of spots detected. The range of spot intensities varied from 64–19342 ppm (300-fold).

TABLE 5.1 A comparison of the features of 2DE reference maps generated by silver staining of total protein and pulse labelling of newly synthesised proteins with radioactive methionine.

Silver stained pulse labelled D=0.02 D=0.03 number of gels analysed1 45 number of independent cultures 3 2 Total features detected ~1800 ~1600 Total matched spots2 1020 870 Dynamic range3 45-fold (220-9881) 180-fold (109-19342) RMG file 026RMGss.tif 030RMGs35.tif

1. number of gels used to compose the Raw Master Gel (RMG) 2. Number of spots matched in at least three gels 3. Dynamic range is defined as the fold difference between the two detected spots with the highest and lowest intensities.

5.3.2 Comparison of silver stained and radiolabelled 2DE images

The main aim for employing the pulse labelling method was to increase the dynamic range of detected proteins while increasing the sensitivity and resolution due to decreased sample loads (10-15 g vs 75 g protein for silver stained gels). In this study dynamic range was on the order of 300-fold for radiolabelled gels compared with 45-fold for silver stained gels. This range is lower than the 1,000 fold previously reported for MelanieII (BioRad) analysed radiolabelled gels (Fegatella et al., 1999). The difference is a result of differences in spot quantitation algorithms employed by each application. For example, the MelanieII analysis software package 130 Chapter 5

35 (Biorad) reports the absolute quantity of radiolabel(counts) present in [ S] L- methionine labelled spots, while Z3 uses a normalised and calibrated measure of relative abundance (ppm). Despite an increase in dynamic range the resolution of radiolabelled gels was lower than for silver stained gels.

Qualitative differences in patterns were also noted between radiolabelled and non- labelled gels. In contrast to silver stained gels, 26.3 % of the total amount of 35S-Met was incorporated into just ten proteins observed on gel images. Nine of the twenty most abundant proteins form a distinct cluster in the 70-72 kDa, pI 4.9-5.0 range. A number of spots from this cluster were noted previously by Fegatella and Cavicchioli

(2000) in a study of gene expression during starvation. The expression of these proteins, referred to as the S1-S5 cluster, is upregulated by more than 10-fold after

24 h of starvation relative to exponential growth phase in ASWG. Although they display a high intensity on radiolabelled gels of starved (Fegatella, 2002) and slowly growing cultures (this study), the relative intensity of these spots on silver stained gels is much lower (Fig 5.4). This is also true for a number of other protein spots that are detected by pulse labelling but are invisible on silver stained gels, eg. S6, S7 and

S8, (Fegatella, 2002) and vice versa, eg m118, m044 (Fig. 5.2)

5.3.3 Possible explanations for quantitative differences between methods

The quantitative differences of corresponding protein spots observed on silver stained and radiolabelled gels of the same condition may be explained by a number of factors. Those proteins that appear to be highly expressed on radiolabelled gels but not silver-stained gels may be a subset induced in response to the added

131 Chapter 5

methionine/cysteine labelling mix, although this is unlikely given the short labelling window (1 h). Mixed amino acids are likely an important source of nutrients for S. alasskensis and other marine heterotrophs and phototrophs (Coffin et al., 1989,

Zubkov et al., 2003). S. alaskensis has a demonstrated constitutive high affinity uptake system for amino acids which has an affinity for alanine higher than any previously reported (Schut, 1993). Methionine is likely to be transported by the same transport system, which has a half saturation constant (Kt) for amino acid uptake of

35 less than 2 μM (Schut, 1993). The final concentration of [ S] L-methionine added during labelling was approximately 20 nM. Most radioactivity was incorporated into insoluble material within the first five minutes and more than 66% of total radioactivity was incorporated during the 60 min labelling period (Fig 5.2). The capacity to transport and incorporate methionine efficiently, and the high level of total incorporation indicates that there were no inherent problems with labelling the chemostat cultures.

The different distribution of incorporation into protein spots may also reflect differences in proportion of methionine and cysteine residues contained within each protein. That is, methionine and cysteine rich proteins will display a relatively higher signal than proteins of similar abundance and molecular weight which have fewer methionine and cysteine residues. A further explanation may be drawn from the dynamics of silver staining, as the intensity of spot staining can be influenced by post translational modifications, localised staining artefacts and the amino acid composition of the protein (Merril et al., 1984). Finally, unequal rates of synthesis and degradation for a specific proteins may account for the differences observed.

Differential synthesis and degradation rates have been recorded for a range of 132 Chapter 5

bacterial proteins. A relevant example is the E. coli RpoS sigma factor where rpoS is continually transcribed but the level of mature protein is regulated by proteolysis by

ClpXP (Schweder et al., 1996, Zgurskaya et al., 1997). Further evidence for different protein turn-over rates in bacteria comes from analysis of protein expression throughout the cell cycle of (Grünenfelder et al., 2001). In this organism a large number of proteins were found to be synthesised and degraded rapidly in a cell-cycle dependent manner. In contrast to synchronised C. crescentus cultures in the previous example, asynchronous steady state chemostat cultures are composed of a heterogeneous population of cells with respect to their position the cell division cycle. As a result, the total number of proteins observed on non-labelled

2DE gels represent an average of all proteins expressed throughout the growth and division cycle of the organism under the chemostat defined conditions. That is, the abundance of each protein reflects the relative abundance within an average cell.

Despite this, during pulse labelling, the short labelling window used may produce a bias for proteins that have higher synthesis rates but lower overall abundance relative to other proteins. In other words, proteins that are synthesised at a faster rate will be preferentially labelled and their relative abundance in radiolabelled vs silver stained gels will depend on their rate of degradation relative to the labelling window

133 Chapter 5

pI pI

Mr (kDa)

FIGURE 5.2 Comparison of [35S] – L methionine labelled and silver stained 2DE protein reference maps for S. alaskensis RB2256. Images were acquired from (A) densitometry of silver –stained gels or (B) phosphor screen of steady state glucose limited chemostat cultures of S. alaskensis RB2256 at = 0.03 h-1. Linear IPG pI range 4-7 and 11.5% polyacrylamide. Molecular weight was determined by comparison with Broad Range Molecular Weight Markers (Biorad).

134 Chapter 5

S1-S5 cluster

132 A 132 130 AB130

FIGURE 5.3. Distinct features present in radiolabelled and silver stained gel images. A subsection of the gel images in figure 5.2 have been enlarged and annotated. The position of the S1-S5 cluster is shown on a representative radiolabelled (A) and silver stained gel (B). The relative position of landmark protein spots 130 and 132 are also presented.

5.3.4 Representation of the S. alaskensis proteome in the pH 4-7, Mr 10-200

kDa window

Based on the analysis of a range of bacterial , with a genome of 3.2 Mbp

(Fegatella, 2002) and an estimate of one gene per kilobase of DNA, S. alaskensis is predicted to produce a total complement of ~3,200 proteins. This estimate is in line with the prediction of 3,869 open reading frames (ORFs) from 3.8 Mbp and 2,515

ORFs from 2.4 Mbp for Novosphingobium aromaticivorans F199 and

Synechococcus sp. strain WH8102 respectively (www.jgi.doe.gov/microbial). The predicted proteomes of two closely related -proteobacteria display the theoretical

135 Chapter 5

distribution of proteins throughout the pH range from 2 to 14 and the Mr range from

1.0 to 240 kDa (Fig 5.5). More than half of predicted proteins (~52-54%) fall within the pH interval 4 to 7 and the Mr range 10 to 150 kDa. Less than 1% have pI values less than 4 and the remainder are spread evenly from pH 7-13.

Broad range immobilised pH gradients (IPGs) were also used to resolve S. alaskensis proteins in the pH 3-10 range (results not shown). However, very few proteins were resolved above pH 7 or below pH 3. The total number of resolved protein spots was also lower than for narrow range IPGs (pH 4-7). Wasinger et al. (2000), and Büttner et al. (2001), have noted a disparity between the number of observed and expected proteins spots on 2DE gels above pH 7 for genitalium and respectively. Analysis of the theoretical proteome of B. subtilis revealed that most of the proteins that were predicted to contain more than one membrane spaning region, i.e. membrane proteins, would be found in the alkaline (> pH 7) region

(Büttner et al., 2001). However, out of the 42 B. subtilis proteins extracted from the alkaline region that have been characterised so far, none of the identified proteins were membrane proteins (Büttner et al., 2001). This observation is in line with the data from a number of laboratories which suggest that membrane proteins are generally not separated and visualised with standard methods (Wilkins et al., 1998,

Santoni et al., 2000).

Taking into account that about 25% of all bacterial proteins are membrane proteins

(Hedman et al., 2002), the 1,020 reproducibly detected spots represent about 50% of the soluble fraction of the S. alaskensis proteome. Based on estimates obtained from the predicted proteomes of closely related -proteobacteria (Fig 5.4) we can predict

136 Chapter 5

that about 50% of all proteins will be present within the pH range 4 to 7. For S. alaskensis the maximum number of spots that can be visualised on a single gel in the pH range from 4 to 7 is 1,800. This is relatively higher than the predicted and actual values for other bacteria and may reflect a largely constitutive pattern of gene expression or a higher rate of post-translational modification in S. alaskensis

(Fegatella, 2000). In fact, studies with other organisms show that the level of post- translational modification can be significant in bacteria, thus increasing the number of potential protein products per gene. Estimates of the number of protein products per gene in the proteomes of E. coli, Pseudomonas aeruginosa and M. genitalium range from 1.22 to as high as 1.7 (Nouwens et al., 1999, 2000, Wasinger et al.,

2000). The trains of protein spots observed on S. alaskensis gels in the high molecular weight range with differing pI are indicative of post-translational modifications (Fig 5.3). It is noteworthy that similar chains of spots are observed for high molecular weight proteins in the B. subtilis proteome. Some of these spots have been characterised and correspond to EF-Tu, EF-G, AhpC, AhpF and GroEL

(Büttner et al., 2001).

137 Chapter 5

A

B

Figure 5.4 Theoretical 2DE protein maps for -proteobacteria closely related to S. alaskensis. A. theoretical 2DE protein map of Caulobacter crescentus,B.theoretical

2DE protein map of Novosphingobium aromaticivorans F199. Theoretical 2DE protein maps were compiled from predicted pI and molecular weight data generated from the complete genome sequence of C. crescentus and the draft genome sequence of N. aromaticivorans F199. The pI 4-7, 5,000 – 150,000 Da window is marked on each.

138 Chapter 5

5.3.5 Conclusion

To my knowledge this is the first demonstration of [35S] L-methionine/cysteine pulse labelling applied to any form of chemostat culture. In terms of dynamic range and sensitivity, pulse labelled gels present a distinct advantage over silver stained gels for analysis of protein expression, while silver stained gels display a higher resolution.

The reference images of S. alaskensis demonstrate that 2DE gels of proteins in the pH interval 4-7 can be used to examine the expression of a large proportion of soluble, abundant, cytosolic proteins. Furthermore, pulse labelling of newly synthesised proteins and silver staining of total proteins are complementary methods for analysing gene expression in S. alaskensis.

139 Chapter 6

6EXAMINATION OF GENE EXPRESSION IN

RESPONSE TO NUTRIENT LIMITATION IN S.

ALASKENSIS

140 Chapter 6

6.1 BACKGROUND

The use of proteomics to examine global changes in gene expression is no longer novel.

Numerous studies have used proteomics successfully to monitor changes in protein profiles in response to defined environmental conditions and controlled perturbations.

A few examples that demonstrate the numbers of proteins spots that can be simultaneously analysed are provided in table 6.1. Over the past decade significant advances have been made in areas of protein preparation and fractionation for 2DE separations and also in the available software for analysing increasing numbers of replicates and resolved spots. Until recently a major limitation of the 2DE approach has been the reliance on N-terminal sequencing, also known as Edmann-degradation, for the identification protein spots for gene identification, utilising reverse genetics involving

PCR with degenerate primers designed from sequence tag information. N-terminal sequencing is expensive and requires large amounts of starting material while reverse genetics is time consuming and not trivial and therefore a major limiting factor on the number of spots that can be identified.

The exponential increase in genome sequence projects over the past decade has allowed for the application of mass spectrometric methods for direct spot identification, by comparing peptide mass fingerprints obtained by time-of-flight (TOF) mass spectrometry (MS) with peptide masses theoretically derived from sequences in databases. Peptide mass fingerprinting (PMF) is made possible by the high degree of accuracy of mass spectrometers and the uniqueness of peptide masses derived from proteolytic digests of pure proteins, i.e. each peptide mass can only be represented by a limited number of amino acid compositions. Thus, confident identifications can be

141 Chapter 6

made with as few as two or four peptide masses and as little as 10 pg of starting material in comparison with tens of gs of material for N-terminal sequencing. One drawback of PMF is that, in order to identify a spot taken from a gel, the gene sequence for that spot needs to already be available in the database. Thus PMF is largely limited to organisms with significant amounts of their genome sequenced.

Despite this drawback protein homologues in different species tend to have conserved primary sequences, raising the possibility of cross species database matching. In practice only extremely well conserved proteins can be identified by this method because a change of a single amino acid within the span of a peptide of any length is enough to change the peptide mass (Cordwell et al., 1995, Cordwell et al., 1997). A potential fix to this problem is to utilise a number of other parameters that are conserved across species boundaries, such as amino acid composition, pI, and Mr (Table 6.2) in conjunction with PMF and specialised database searching software for spot identification, MultiIdent (Wilkins et al., 1998). The benefits of cross species matching in conjunction with PMF are obvious. For example Oosthuizen et al. (2002) observed

24 out of 345, or 7% proteins unique to one of two periods during biofilm formation by

Bacillus cereus cells. Thirty four out of 345, or nearly 10% of the total proteins detected were upregulated. Of the numerous differences detected only 8 of these proteins were present in enough quantity for identification by N-terminal microsequencing and the remainder were unidentified. In comparison, Weekes et al.

(1999) managed to separate 1125 protein species from bovine ventricular tissue. Of 35 proteins differentially expressed in diseased compared to normal tissue 12 could be identified with the combined data of amino acid composition, peptide mass fingerprinting and N-terminal microsequencing in combination with MultiIdent.

142 Chapter 6

TABLE 6.1. Examples of the number of unique, differentially expressed and total proteins analysed in selected bacterial 2DE projects.

Organism Experiment D1 T2 Reference

C. crescentus Cell cycle 144 979 Grünenfelder et al., progression 2001 P. angustum Carbon 301 > 1000 Nyström et al., 1992 S14 starvation E. coli Phosphate 413 816 Vanbogelen et al., restriction 1996 S. alaskensis Glucose 249 1,086 Fegatella and starvation Cavicchioli, 2000 Pseudomonas Osmotic stress 132 950 Vasseur et al., 1999 fragi Bacillus cereus Biofilm 24 345 Oosthuizen et al., 2002 formation 1. Number of differentially expressed proteins 2. Total number of proteins analysed

During the course of this study another mass spectrometric tool became available, known as micro-liquid chromatography electrospray ionisation tandem MS (LC-ESI

MS/MS). A schematic of LC-ESI MS/MS is given in Figure 6.1. Tandem MS is similar in principle to MS, once distinct peptide masses are obtained from the first TOF-

MS, a discrete peptide ion is selected, fragmented by collision with an inert gas

(collision induced dissociation, CID) and the daughter ions subjected to another round of TOF-MS. Since the peptides tend to fragment at their peptide bonds, spectra generated from a population of daughter ions can be used to interpret the sequence of the parent peptide. Because this method produces peptide masses as well as associated sequence tags it is more accurate and less restricted than PMF, which requires exact mass matches, for cross-species identification. LC-ESI MS/MS has the added benefit of a preliminary chromatography separation of the peptides generated by proteolysis, which is an advantage for identifying spots that often contain multiple protein species.

Sequence tags can be generated de novo by manual interpretation of the tandem mass 143 Chapter 6

spectra, however LC-ESI MS/MS produces large amounts of complex CID spectra from a single protein (up to 1000). Software such as SEQUEST (Eng et al., 1997) can be used to correlate uninterpreted CID mass spectra with theoretically generated spectra from sequence databases.

The aim of this study is to define and characterise the molecular basis of the response to general nutrient limitation and its links with stress resistance in S. alaskensis RB2256.

Two dimensional gels were employed to analyse global gene expression at different nutrient-limited rates of growth. Complementary methods for producing high-resolution

2DE gels from total protein and pulse-labelled chemostat cultures were described in the previous chapter. In the absence of genomic sequence information cross species database matching was used with data obtained from N-terminal sequencing, PMF, amino acid analysis and sequence tags obtained from tandem mass spectrometry for spot identification. The identification success rates of various methods were compared, the results of spot identification demonstrate the importance of general stress response genes during slow nutrient limited growth.

144 Chapter 6

  Separate by Excise spot LCto 2DE  In-gel digest database separate  Analyse profiles (trypsin) ESIMS MS search  Define  Extract peptides peptides

2nd Quadrupole 1st Quadrupole (Collision 3rd Quadrupole

SEQUEST Fragmen MultiIdent t Mowse Prowl Peptides Peptide Mass Peptide sequence

2DE and analysis

Figure 6.1. Outline of protein spot identification strategy using micro Liquid Chromatography Electrospray Ionisation Tandem Mass Spectrometry( LC ESI MS/MS) and a triple ‘quadrupole’ mass spectrometer. Protein spots of interest are excised from 2DE gels and digested in-gel with trypsin. The resulting peptides are extracted from the gel, separated by an in-line LC, injected in the tandem mass spectrometer by electrospay ionisation. In the first quadrupole distinct mass ion peaks are recorded and selected for fragmentation by collision with an inert gas in the second quadrupole. Peptide sequences can be determined by MS of the generated fragments in the third quadrupole. The combined data of peptide mass fingerprint and peptide sequences as well as pI and Mr generated from 2DE are then used to search sequence databases with SEQUEST and related software.

145 Chapter 6

6.2 MATERIALS AND METHODS

6.2.1 Two-dimensional electrophoresis, image acquisition and analysis

Sample preparation, two-dimensional electrophoresis, image acquisition and analysis were carried out as described in previous chapters unless stated below.

6.2.2 Preparative 2DE and semi-dry blotting

Preparative 2DE was carried out as for analytical 2DE (Chapter 5) with the exception that 200 g to 2000 g of total protein was loaded onto the IPG strip. Electroblotting of proteins from 2DE polyacrylamide gels was carried out according to a method modified from Dunn (1999) with a BioRad Trans-Blot SD Semi-Dry Transfer Cell. After SDS-PAGE, gels for electroblotting were immediately washed in MilliQ (2 x 10 min) and then soaked in cathode buffer (0.04 M 6-amino-n- hexanoic acid, 0.025 M tris, 20% v/v methanol, 2 x 5 min). Prior to completion of SDS-PAGE twelve pieces of filter paper (Whatman 3MM, cut to the dimensions of the gel) were washed in MilliQ (3 x 5 min) and drained. Half of the washed filter paper was soaked in cathode buffer and the other half soaked in anode buffer (0.3M tris, 30% v/v methanol) until the blot was assembled. Polyvinylidine diflouridine (PVDF) membrane was pre-wetted in methanol and washed in cathode buffer (2 x 5 min). The blot was assembled from the bottom as follows: 6 pieces of filter paper soaked in anode buffer were placed on the anode followed by the sheet of PVDF membrane, the gel and then the remaining 6 pieces of filter paper soaked in cathode buffer. The buffer-soaked pieces of filter paper were blotted briefly prior to assembly and each layer of the blot was carefully rolled with a glass rod to remove air bubbles. The cathode lid was closed over the assembly and the blot carried out at 0.8 mA cm-2 for 1.5-2.0 h in a 4˚C room.

6.2.3 Blot staining

At the completion of the blot the membrane was washed in MilliQ (2 x 5 min), stained with freshly prepared 0.25% (v/v) coomassie brilliant blue R250 in 40% (v/v)

146 Chapter 6

methanol for 2-5 min, destained in 30% (v/v) methanol for 5 min and then allowed to dry. Dry PVDF blots were stored in plastic zip-lock bags at 4˚C.

6.2.4 Sample preparation and in-gel trypsin digestion

For matrix assisted laser desorption/ionisation time of flight mass spectrometry (MALDI-TOF-MS) samples were prepared from spots on PVDF membranes. All work areas, utensils and tubes were rinsed with a 50 % (v/v) methanol, tri-fluoro acetic acid (TFA, 0.1 % v/v) solution and allowed to air dry. Spots were excised from PVDF membranes with a sterile razor blade, sliced into 1 mm2 segments and placed in 0.5 ml microfuge tubes. After rehydration in 50 % (v/v) methanol the membrane pieces were dehydrated in 200 l of acetonitrile by vortexing for 30 s. The solvent was removed and the membrane pieces completely dried under vacuum in a DNA 120 SpeedVac (Savant Instruments). Sequencing grade, methylated, porcine trypsin (Promega) was added (20 – 50 ng total) to a sufficient quantity (3-10

l) of digestion buffer (25 mM NH4CO3, 1 % w/v octyl--glucoside, 10 % v/v methanol) to cover the membrane pieces. After overnight digestion at 28˚C the aqueous portion containing the extracted peptides was removed and stored in a fresh tube. The remaining peptides were extracted from the membrane pieces by sonication for 30 min in 10 l of freshly prepared 50 % (v/v) ethanol and 50 % (v/v) formic acid. The extractions were combined, dried under vacuum and rehydrated immediately prior to mass spectrometry in 1 to 5 l of 10 % (v/v) acetonitrile and 1 % (v/v) formic acid.

For micro-capillary liquid-chromatography coupled with an electron-spray ionization mass spectrometer (LC-ESIMS) analysis, protein spots of interest were directly excised from silver stained gels and preparative coomassie g250 stained gels. To reduce contamination from human keratin and dusts, all utensils were wiped with methanol just prior to use. Gels were placed in a methanol cleaned Pyrex container and protein spots excised with a sterile razor blade. BSA was excised and served as positive control for digestion. For each protein sample, spots were excised from 2 to 4 gels, pooled in a 0.5 ml microfuge tube and dehydrated in acetonitrile at room temperature for 10 min. The gel pieces were dried under vacuum and rehydrated with

147 Chapter 6

no more than 10 μl trypsin (12.5 ng l-1), 50 mM ammonium bicarbonate (pH 8.0) at 4oC for 45 min. Following rehydration, excess liquid was removed and digestion performed at 37oC overnight. The peptide solution was collected and the gel pieces further extracted with 20 mM ammonium bicarbonate followed by 3 extractions with 5 % formic acid, 50 % acetonitrile for 20 min each at room temperature. All samples were dried under vacuum to a volume of approximately 5 l and stored at -20oC.

6.2.5 Mass spectrometry analysis

Peptide mass fingerprinting (PMF) was carried out by MALDI-TOF-MS with a Voyager DE STR mass spectrometer. Briefly, peptide mixtures from tryptic digests were mixed with an excess of matrix (-cyano-4-hydroxycinnamic acid [CHCA]) in 80 % acetonitrile on the target plate and peptide mass fingerprints obtained. Substance P (Sigma) was used as an internal calibration standard, and the number of laser shots accumulated were 50-100. Mass spectra were analysed with PerSeptive Biosystems Data Explorer™ software version 3.2. The spectra from target samples were compared with the spectrum from BSA, and a section of the membrane without any stained proteins. Background peaks, corresponding to gel and matrix components, were subtracted to produce peptide mass fingerprints for each protein.

Sequence tag analysis by LC-ESI MS was performed by loading 2-4 l from 5 l of peptide sample (<10 pmol) onto a μ-capillary C18 200dia, 5 μm(Michrom BioResources, USA) bulk packaging column. Eight cm length reverse phase columns were prepared in house using a pressure vessel (Brechbühler Inc., Switzerland). Peptides were eluted over a 30 min gradient (Acetonitrile 80 %, acetic acid 2 % and HFBA 0.004 %) using an HP1100 with T-splitter to give a flow rate of 300 nl min-1. Tandem mass spectrometry was performed with a TSQ7000 (Thermofinnigan, USA).

All CID of peptides above an ion count threshold of 5000 were monitored. ESI voltage was set to 1.3 kV and capillary temperature of 190°C was used. Proteins were identified using SEQUEST (Thermofinnigan, USA) and all identifications were verified manually by database cross-checks. SEQUEST parameters were defined as:

148 Chapter 6

no enzyme; deviation of parent ion mass <3.5 Da; no modifications. Matches were only considered if fragments were tryptic, the Xcorr score was >2 for MH2+/2, and a distinct ladder sequence was observable. These criteria were used to determine confident CID spectra and identifications. The MS/MS spectrum was analysed using SEQUEST software (Corthals et al., 2000) to search a local database of Novosphingobium aromaticivorans F199 translated sequences obtained from http://www.jgi.doe.gov/JGI_microbial/html/ and the NCBI non-redundant (nr) database (ftp://ftp.ncbi.nih.gov/blast/db/).

6.2.6 Amino acid and Edman degradation

For spot M044, amino acid analysis was performed at the Australian Proteome Analysis Facility (Macquarie University, Sydney) and N-terminal sequencing by Edman degradation at the Biomolecular Resource Facility (Australian National University, Canberra).

149 Chapter 6

6.3 RESULTS

6.3.1 Differential expression of proteins observed on silver stained gels

To determine the changes in gene expression that are associated with high and low rates of growth, and to identify the genes involved in the increased levels of hydrogen peroxide resistance at low rates of growth, 2DE profiles of total cellular protein were generated for cells grown at 0.03 h-1 and 0.18 h-1 (Fig. 6.2). Raw master gel images were generated from four replicate gels of 0.026 h-1 and five replicate gels of 0.18 h-1 (Table 6.2). Protein spots from each condition were separated evenly throughout the resolved pI range from 4 to 7 and molecular mass from 5 to 150 kDa. Up to 1800 protein spots were detected on each silver-stained gel and 1,023 spots conserved from a majority of gels of each condition were analysed.

A small number of differences were observed between the protein profiles of gels from different rates of growth. Spots that are differentially expressed by more than 2- fold at one growth condition compared to another are highlighted in Figure 6.2, and the total number of spots specific for each growth rate is summarised in Table 6.2.

This table shows that only 20 spots (~1.7 %) differ significantly between the gels: 10 specific to a low rate of growth (0.03 h-1) and 10 specific to a high rate of growth

(0.18 h-1). Each spot marked in Figure 6.2 was characterised for its approximate molecular weight, isoelectric point and difference in relative spot intensity (Table

6.2).

150 Chapter 6

The differentially expressed protein spots vary in predicted molecular weight from

20.5 to 63 kDa and predicted isoelectric point from 4.2 to 6.3. The largest difference in intensity for a spot that was present in both master gels was for spot M142 that is

5.9-fold more intense at growth rate 0.18 h-1 compared with 0.03 h-1. Spots M433,

M022, M021, M155 and M001 were non-detectable in one growth condition. Based on the detection limit for the 1023 spots examined (220 ppm), spot M001 had the highest level of intensity (5786 ppm) indicating at least a 26-fold higher level of spot intensity at a growth rate of 0.18 h-1.

TABLE 6.2. Summary of the number of features detected and the number of spots analysed from replicate silver-stained and radiolabelled gel images.

Specific Number of Featuresa number of growth rate replicates analysed spots (h-1) (n) Silver stained 0.03 (4) ~1600 1023 0.18 (5) radiolabelled 0.03 (5) 1400-1600 870 0.18 (4) n: the number of gels used to create master gel images for analysis a: the total number of features detected

151 Chapter 6

152 Chapter 6

TABLE 6.3 Summary of characteristics for differentially expressed protein spots where expression is increased or decreased by more than 2-fold at a high (0.18 h-1) relative to a low (0.03 h-1) on silver stained 2DE gels (Fig 6.2)

1 2 Spot ID DI pI Mr M433 25.0* 4.9 30 Decreased M022 16.7* 5.4 31 at = M118 11.1 4.9 30 0.18 h-1 M074 5.3 4.8 36 () M588 3.8 4.2 26 M141 3.3 5.2 41 M044 3.0 4.7 29 M330 2.6 5.0 58 M117 2.5 4.7 37 M114 2.5 4.5 21 Increased M001 26.3* 4.8 54 at = M155 10.5* 5.1 43 0.18 h-1 M021 10.5* 5.4 32 () M142 5.9 4.6 54 M037 4.1 4.5 32 M156 3.6 5.1 43 M166 3.0 5.0 41 U1 2.9 5.7 52 M323 2.2 6.3 27 M132 2.1 5.1 63

1. Where possible spot IDs are kept consistent with previously published work 2. Differential spot intensity is expressed as the fold increase or decrease at =0.18 h- 1 relative to 0.03 h-1 * not detected on all gels of the opposite condition, differential expression was calculated relative to the limit of detection for those gels.

153 Chapter 6

Figure 6.3. Differential intensity for protein spots from S. alaskensis cultures grown at low relative to high rates of growth in glucose-limited chemostats. Protein spots were visualised by silver-staining and quantified with Z3 Software. Spots with positive DI values have higher levels of expression at low relative to high rates of growth. Error bars represent standard deviations of the mean values.

154 Chapter 6

6.3.2 Differential expression of pulse labelled proteins

In light of the few differences that were observed between protein profiles generated from total protein at high and low rates of growth the alternative method of radioactive pulse labelling was employed. It was reasoned that the superior dynamic range of radiolabelled gels over silver stained gels may reveal subtle differences in expression. Pulse labelling with 35S-Met/Cys was conducted for chemostat cultures growing at = 0.03 and 0.18 h-1. Labelling was carried out in small scale chemostats

(50 ml) over a period of 1 h. The initial rate of incorporation of 35S Met/Cys was similar for cultures growing at high and low rates of growth (Chapter 5). Raw master gel images were generated from five replicate gels of 0.03 h-1 and four replicate gels of 0.18 h-1 (Table 6.2). Biological replicates were generated from two independent chemostat cultures for each growth rate.

More than 1,400 features could be detected on each gel image and 752 spots conserved from a majority of gels of each condition were analysed (Figure 6.4). A total of 102 protein spots were found to be differentially expressed by more than 2 fold at high or low rates of growth. The most significant 32 spots (~ 4.2% of the total) were differentially expressed by more than 2.5 fold: 19 were specific to a low rate of growth and 13 were specific to a high rate of growth (Fig. 6.5). Each spot marked in Figure 6.4 was characterised for its approximate molecular weight, isoelectric point and difference in relative spot intensity (Table 6.4). At =0.03 the

12 spots in the S1-S5 cluster (Chapter 5 and Fegatella and Cavicchioli, 2001) account for more than 9% of the total radioactivity across the entire gel image.

155 Chapter 6

A summary of the quantitative differences between protein spots at low and high rates of growth from both silver stained and radiolabelled analyses is given in Table

6.5. It is noteworthy that the majority of differentially expressed spots identified by each method were different (Tables 6.3 and 6.4). Out of a total of 17 spots that were specific to a high rate of growth, five spots were common to both analyses (M001,

M021, M037, M152 and M155). Out of a total of 26 spots that were specific to a low rate of growth only three spots were common (M022, M118 and M074).

156 Chapter 6

157 Chapter 6

TABLE 6.4. Summary of characteristics for differentially expressed protein spots where expression is increased or decreased by more than 2.5-fold at a high (0.18 h- 1) relative to a low (0.03 h-1) on radiolabelled 2DE gels (Fig 6.4).

1 2 Spot ID DI pIMr (kDa) Decreased S3 156* 5.0 72 at = S1 140* 5.0 71 0.18 h-1 M1040 116* 28 29 () M1046 109* 4.3 50 S5 103* 5.0 69 M074 85 4.8 36 M1106 50* 5.1 33 M1262 49 6.5 35 S4 41 4.9 69 L3 32 5.9 33 S2 31 4.9 74 M1332 21 4.5 29 M1410 18 4.5 29 M1630 4.5 5.3 30 M1478 4.4 5.4 70 M1054 4.2 5.9 36 M022 4.1 5.4 31 M1196 3.0 6.8 29 M118 2.8 4.8 29 Increased M1585 54* 5.2 50 at = M037 46 4.7 33 0.18 h-1 M1586 31 5.3 50 () M001 12 4.6 54 M155 11* 5.1 43 U2 10 5.5 52 M2032 6.9 6.1 38 M021 6.2 5.4 32 M1422 5.4 6.2 31 M1274 5.2 4.4 27 M1150 3.6 6.5 33 M1094 3.3 4.8 62 U1 3.0 5.7 52

1. Where possible spot IDs are kept consistent with previously published work 2. Differential expression is expressed as the fold increase or decrease at =0.18 h-1 relative to 0.03 h-1 * not detected on all gels of the opposite condition, differential expression was calculated relative to the limit of detection for those gels.

158 Chapter 6

Figure 6.5. Differential intensity for radiolabelled protein spots from S. alaskensis cultures grown at low or high rates of growth in glucose-limited chemostats. Radiolabelled protein spots were analysed and quantified with Z3 Software. Error bars represent standard deviations of the mean values.

159 Chapter 6

TABLE 6.5. Summary of differentially expressed protein spots on silver stained and radiolabelled 2DE gels.

silver stained Criteria Decreased Increased at high at high 2–4fold 6 5 5–9fold 0 2 >10fold 3 3 Total 9 10

Pulse labelled Criteria Decreased Increased at high at high 2 – 2.4 fold 22 48 2.5 – 9 fold 6 7 > 10 fold 1 3 >20fold 12 2 Total 19 12

6.3.3 Identification of protein spots on the basis of peptide mass fingerprinting and Edman sequencing

To obtain pure protein in sufficient quantities for characterisation, preparative quantities of total protein (2.0 – 3.0 mg) from steady-state chemostat cultures of S. alaskensis were separated by 2DE (pH range 4-7) and electroblotted to polyvinylidene difluoride (PVDF) membrane. Electroblotted protein was visualised by staining with coomassie brilliant blue R250. Variable results were obtained with repeated blots. In general the sensitivity and resolution of coomassie-stained PVDF blots was less than silver-stained analytical-gel equivalents (e.g. Fig 6.2).

Approximately half of the total proteins resolved and detected on silver-stained analytical gels could be detected on coomassie-stained PVDF blots. As a

160 Chapter 6

consequence, a fraction of proteins with altered expression levels from Table 6.3 have either not been visualised (M141, M142), cannot be clearly resolved (M155,

M118) or are barely visible (M001, M037, M117).

Spots M021, M044, M141 and M114 were digested with trypsin and the resulting peptides were analysed by MALDI-TOF MS. The obtained spectra were compared with the spectrum obtained from a control sample excised from the same piece of membrane. Background peaks, corresponding to gel and matrix components, were subtracted to produce peptide mass fingerprints for each protein. Mass spectra for spots M021 and M141 were not obtained due to low signal to noise ratio. Amino acid composition data was obtained for spot m044 from the Australian Proteome analysis facility (APAF).

Peptide mass and amino acid composition data for spot M044 were used to search the SwissProt and TrEMBL protein databases using the MultiIdent tool (Wilkins et al., 1998). Seven of the top ten ranked integrated scores correspond to bacterial chaperonin GroEL homologues or the closely related Rubisco subunit binding- protein  (integrated scores 8.7-13.5). On the basis of this result spot M044 was putatively identified as the chaperonin GroEL.

To confirm this identification spot M044 was excised from a duplicate blot and subjected to N-terminal sequencing. The first 16 residues of M044 were determined by Edman degradation (AAVKAPGFGDRRKAML) and used to search the Swiss-

Prot protein sequence database using BLAST. The top 99 matches were 100 % identical to a region that corresponds to residues 274 – 290 of the E. coli GroEL, 161 Chapter 6

thereby confirming its identity (Table 6.6). The molecular mass and pI of M044 is consistent with the predicted Mr and pI of the C-terminal portion of E. coli GroEL.

6.3.4 Identification of protein spots by LC-ESI MS/MS

Collision-induced dissociation (CID) spectra were obtained for U1, M001, M114,

M115, M117, M118, M130, M141, M156, M166, M191 and M433. On the basis of

SEQUEST searches confident identifications were made for U1, M001, M115,

M117, M130, M156, M166 and M433 (Table 6.6). In all cases the best correlation scores were to predicted coding DNA sequence translations (CDS) from the N. aromaticivorans genome sequence, and in some cases equally good matches were obtained to other bacterial sequences

U1 matched the ATPase  subunit. U1 has a predicted Mrof52,000Dawitha predicted pI of 5.9, consistent with the predicted mass of ATPase  subunits in other species. M001 matched the ATPase  subunit. The pI was 4.8 and Mr 48,000 for

M001, again, consistent with the Mr of this protein in other species. Spot M044 was determined by N-terminal Edman sequencing to be the C-terminal portion of GroEL

(Mr 29,000). M130 was also found to be GroEL, and its Mr is consistent with a full- length E. coli protein (58,000). Spot M156 produced good matches to DNA polymerase III -subunit and EF-Tu. Matches to two unrelated proteins indicated this spot was likely to have resulted from the co-migration of two or more proteins.

Similar occurrences of co-migrating proteins have been found in a recent evaluation of 2DE gel based techniques with S. cerevisiae and E. coli (Nouwens et al., 1999).

The Mr of M156 (43,000 Da) is comparable to E. coli EF-Tu (42,000 Da) and DNA 162 Chapter 6

polymerase III -subunit (41,000 Da). In addition to M156, spot M115 (42,000 Da) matched EF-Tu and was consistent with the Mr of E. coli EF-Tu (42,000Da). Both

M156 and M115 have the same pI and only differ marginally in Mr, however M115 is a more abundant protein (Fig. 6.2). Single matches to an electron transfer flavoprotein  subunit, a NADPH-sulphite reductase  subunit and succinyl-CoA synthetase -chain were identified for M433, M117 and M166, respectively.

Using SEQUEST, matches were not obtained for U2, M005, M114, M118, M141 and M191. Sequence tags (peptides 2-5 amino acids in length) were manually derived from the CID spectra, however no confident identifications were obtained

(Table 6.6).

163 Chapter 6

TABLE 6.6. Proteins characterised by mass spectrometry or Edman sequencing from the S. alaskensis proteome.

Spot DI pIMr SEQUENCE Protein Product Xcorr ID 35S silver Invariant M115 1.0 1.0 5.1 42 (K)LLDQGEAGDNIGALIR EF-Tu, Protein 4.50 Translation Elongation Factor Tu M130 1.0 1.0 5.0 58 (R)AAVEEGIVPGGGTALLYATK GroEL, 60 kDa 4.23 AAGDGTTTATVLAQAIVR chaperonin 4.09 (K)VGGATEVEVK 3.76 (R)VDDALHATR 3.48 (R)GYLSPYFITNPEK 3.28 (K)EGVITVEEAK 2.87 (K)AAGVIDPTK 2.52

Increased M433 <2.5 25.0 4.9 30 (-)TFTRPIYAGNAIATVESSDAK EtfA, Electron 3.97 at a low transfer flavoprotein rate of  subunit growth M044 <2.5 3.0 4.7 29 AAVKAPGFGDRRKAML GroEL, 60 kDa NA chaperonin (isoform/fragment)

M117 <2.5 2.5 4.7 37 (K)HILTLGLEPPAAELERIK CysI, NADPH- 2.39 Sulphite reductase  subunit Increased M001 12 26.3 4.8 48 (R)FTQAGSEVSALLGR AtpB, ATP synthase 2.50 at a high  subunit rate of M156 <2.5 3.6 5.1 43 (K)LLDQGEAGDNIGALIR EF-Tu, Protein 4.57 growth (K)TTLTAAITK Translation 2.36 Elongation Factor Tu 3.22 2.69 2.53 (R)NTIPILSNVLIEAAPDSTVR DnaN, DNA (R)FAISTEETR polymerase-III  (K)LIDGTFPDYSR subunit M166 <2.5 3.0 5.0 41 MNVHEYQAK SucC, Succinyl-CoA 3.19 (K)QLPGPLYVVK synthetase  chain 2.34 (R)LEGTNVQQVK 2.75 (K)EVNLSVPLVVR 2.75

U11 3.0 2.9 5.7 52 (K)TAVAIDTFNQK AtpD, ATP synthase 3.88 (R)TGTIVDVPVGK  subunit 3.64 (R)STVAQIVR 2.3 No matches with PMF, SEQUEST or sequence U2, M005, M114, M118, M141, M191 tags

1Spot U1 was present during logarithmic growth but not during starvation (Fegatella and Cavicchioli, 2000) and was present at increased levels at a high rate of growth. NA: Data not available or not relevant.

164 Chapter 6

6.4 DISCUSSION

6.4.1 Differential expression of proteins in steady state chemostats

A rapid means of extensively surveying the gene products required for differential stress resistance is through the analysis of the 2DE protein profiles. Interestingly, there are a relatively small number of changes between the profiles for fast and slow growing cells on both silver-stained and radiolabelled gels (Figs. 6.2, 6.4). This may indicate that the alterations in gene expression are associated with a narrow range of physiological responses, including resistance to hydrogen peroxide (Chapter 4). As the hydrogen peroxide resistance of cells with growth rates of 0.026 and 0.030 h-1 are equivalent and higher than for cells grown at 0.175 and 0.18 h-1, the candidate spots that represent proteins that may be particularly important for the stress resistance state of the cell, are those proteins that are decreased or increased. That is, proteins that are specifically up regulated or repressed at low rates of growth, may either activate or derepress the associated stress resistance mechanism(s). These candidate spots are M118, M117, M141, M115 and M001.

The quantitative differences between the silver-staining and radiolabelling methods were greater than expected. Even though the maximum observed DI was 5-fold greater on radiolabelled gels in comparison to silver-stained gels, the number of differentially expressed spots did not significantly increase. These analyses suggest that growth rate does not induce major changes in gene expression, at least in the soluble fraction of proteins that could be resolved with the methods employed. This may indicate that the dramatic changes in oxidative stress resistance which are linked

165 Chapter 6

to growth rate are mediated through a small number of key changes in gene expression, and/or that additional changes (e.g. metabolites or post translational modifications) may be important. Future work could gain insight into the important gene expression events by examining the changes which occur during a shift from low to high growth rate, and vice versa. The development of radiolabelling in chemostats will facilitate this type of analysis which would largely be based on pulse labelling experiments. Since one might expect a considerable proportion of cellular protein to be associated with the electron transport chain and nutrient uptake capabilities, located within the cytoplasmic membrane, future work to resolve the basis of oxidative stress resistance may also involve analysis of targeted sub-cellular fractions, such as membrane proteins, using established methods (e.g. Nouwens et al., 2002, Huang et al., 2002).

It should be noted that quantitative comparisons of spot number and spot intensity between this study and the analysis of protein expression in response starvation

(Fegatella and Cavicchioli, 2000) are limited due to differences in experimental conditions employed and the analysis software used (Z3 vs MelanieII from BioRad).

In particular, Fegatella and Cavicchioli report up to 70-fold differential spot intensity in an analysis of expression profiles during starvation. This level of dynamic range was not observed in the current study, most likely as a result of the use of the Raw

Master Gel matching of Z3 but possibly also due to the heterogeneity of cells, with respect to the cell division cycle, in steady state chemostat cultures.

166 Chapter 6

6.4.1.1 MultiIdent matching

In the absence of genomic sequence data the identification of proteins in proteome projects relies on multi parameter cross species database matching. The amino acid composition of homologous proteins is highly conserved across species and domain boundaries (Wilkins et al., 1998). While peptide mass data is not highly conserved, it is expected that some peptide sequences, and thus peptide masses, for example, those associated with functional domains, will be conserved (Wilkins et al., 1998 and

Cordwell et al., 1995).

The efficacy of the combined approach to cross species identification has been demonstrated by a number of studies. In a theoretical study MultiIdent was used to successfully identify a group of 10 known proteins from experimentally derived

PMF and amino acid compositional data (Wilkins et al., 1998). In addition to the correct identification each MultiIdent search identified from two to twenty consecutive proteins that were homologues of the target protein in the ‘best integrated score’ list. In another practical example Weekes et al. (1999) have to date successfully identified 12 out of 35 proteins with altered expression levels in a bovine model of cardiomyopathy with MultiIdent.

A MultiIdent search of the SwissProt database using the parameters of PMF and amino acid composition was used to putatively identify spot M044 as a homologue of the bacterial chaperonin GroEL. Seven of the top ten ranked integrated scores correspond to bacterial chaperonin HSP60 homologues and one of the top ten corresponded to the closely related Rubisco subunit binding-protein  (integrated

167 Chapter 6

scores 8.7-13.5). The top 200 high scoring pairs obtained from a Blastp search with the obtained N-terminal sequence were to internal sequence of GroEL homologues from various bacteria. The top 99 matches displayed 100% identity over 16 amino acid residues unequivocally confirming the identity of M044 as a fragment of

GroEL. This result also demonstrates that this portion of the protein is highly conserved amongst diverse bacteria.

The identification of M044 as GroEL provides a positive example of the combined approach to cross-species protein identification in proteomic studies. Presently the success of other identifications is limited by the low abundance of sample material and teething problems associated with obtaining data from samples sent to external facilities for amino acid analysis.

168 Chapter 6

5

S1-S5 Cluster

GroEL m132 ATP Synthase  subunit m142 ATP Synthase  subunit m001

141 U5 RelA m155

Succinyl CoA L3 Synthetase EF-Tu L4 m037

m021 m022

L5 m118 Electron Transfer flavoprotein GroEL1

m588

m114

Figure 6.6. Two dimensional reference image showing the protein spots characterised and their regulation grouping. Low specific growth rate (), and high specific growth rate ().

169 Chapter 6

6.4.2 Cross species identification

PMF analysis has proven useful for protein identification when the genome sequence of the organism is available (Pappin et al., 1997). The identification rate can be considerably lower if the genome sequence is unavailable (Molloy et al., 2001) however in combination with amino acid analysis or other techniques cross species matching is achievable by comparing lists of the top scoring proteins from both techniques and looking for similarities (Liska and Chevchenko, 2003). An important limitation is the incongruity between the evolution of genes and the evolution of the bacterium. For example, a comparison of the complete genomes of E. coli CFT073,

EDL933 and MG1655 revealed that despite being the same species of bacteria, only

39.2 % of the proteins sequences were common to all three strains (Welch et al.,

2002).

N. aromaticivorans and S. alaskensis have recently been reclassified from members of the genus Sphingomonas to the new genera, Novosphingobium and Sphingopyxis, respectively (Chapter 3, Vancanneyt et al., 2001). In contrast to S. alaskensis, N. aromaticivorans is a subterranean bacterium with extensive hydrocarbon degrading abilities (Fredrickrickson et al., 1999). Moreover, preliminary sequence data from S. alaskensis (e.g. stretches of ~8 kb), indicates that the average level of predicted protein sequence identity with N. aromaticivorans is less than 70 % (data not shown). Cross species identification using only PMF data has been reported to be unsuitable when the identity with target proteins is less than 70 % (Lester and

Hubbard, 2002).

170 Chapter 6

Due to these concerns, the primary focus was given to MS/MS spectra, however it was noteworthy that one protein, M044, was successfully identified from the five samples examined using PMF (in addition to N-terminal Edman sequencing). The

CID spectra or derived sequence tags of twenty one samples were matched against protein databases and twelve proteins were identified, providing a success rate of 57

%. In some cases where sequence data was used (e.g. GroEL for spot M044) the amino acid identity was 100 % with peptide sequences from N. aromaticivorans and other bacteria. Many of the proteins are highly conserved across phylogenetic boundaries. For example, the four combined partial peptide sequences from the CID spectra for U3 had identical matches with 98 % of the residues of the ATP synthase

 subunit from N. aromaticivorans and Cytophaga lytica, 94 % with Rhodobacter blasticus and Saccharomyces cerevisiae and93%withhuman. This level of conservation will certainly have assisted the ability to find a cross-species match.

Even with access to the non-redundant protein database the existence of the nearly complete sequence of N. aromaticivorans was a crucial element in the success of cross species matching in that it provided a relatively small, closely related database and therefore a high signal to noise ratio for the matching algorithms. Even though the S. alaskensis genome has now been released (www.jgi.doe.org) the success of this study demonstrates the potential for performing cross species identification with freshly isolated novel environmental strains, or indeed directly identifying proteins in whole community samples taken from the environment using existing genome and environmental sequence data.

171 Chapter 6

6.4.3 Identified proteins

6.4.3.1 Protein synthesis and folding

GroEL was identified with higher relative intensities in proteomes from both high

(M130) and low (M044) growth rates. While M130 appeared to be full length, M044 was found to be the C-terminal portion of GroEL, commencing at a position equivalent to residue 274 in the E. coli protein. GroEL assembles into a protein folding machine by forming an oligomer with two heptameric rings stacked back-to- back, associated with a heptameric ring lid-structure formed by GroES (Chatellier et al., 1998). A minimal GroEL minichaperone comprising residues 193-335 has been reported to exhibit in vitro refolding activity and the ability to complement a temperature sensitive GroEL mutation in vivo (Chatellier et al., 1998). Despite this report, examination of the published crystal structure of the E. coli GroEL-GroES-

(ADP) complex Ala 274 is not found in an area that corresponds to exposed loop region or a junction between two distinct topological or functional domains. This would seem to indicate that M044 is unlikely to have occurred by proteolytic or chemical cleavage of the correctly folded GroEL as part of a multimeric complex that remains functional despite cleavage at Ala 274. It is possible that M044 may have been synthesised by translation initiation at a site internal to the GroEL coding region. The fact that GroEL can form functional minichaperones in E. coli implies that fragments of GroEL can be functional, however the significance of increased levels of the C-terminal GroEL fragment at low growth rates is not known and a protein corresponding to the N-terminal fragment of GroEL is yet to be identified.

172 Chapter 6

The presence of stable fragments of larger proteins has been demonstrated elsewhere

Wasinger and Humphery-Smith (1998) were able to show stable fragments of various proteins that would normally be translated into larger proteins. The presence of these fragments in 2-DE gels can be interpreted as either a true indication of their intracellular abundance and importance, or as experimental artefacts produced during

2DE processing.

EF-Tu is an abundant cellular protein which facilitates the elongation process of translation. The increased abundance of EF-Tu (M156) is consistent with an increased demand for protein synthesis at high rates of growth. In S. alaskensis it has been noted that ribosome synthesis and protein synthesis can be uncoupled in batch cultures (Fegatella et al., 1998, Fegatella and Cavicchioli, 2000), suggesting that ribosomes can perform additional roles, perhaps as nucleotide stores under nutrient replete conditions. Increased abundance of EF-Tu may indicate a tighter coupling of protein synthesis and ribosome levels in nutrient limited cultures as compared with batch cultures.

EF-Tu has also been shown to perform functional roles unrelated to protein synthesis in other organisms (Neidhardt et al., 1990) and it is possible that M156 (as opposed to M115) has an alternative cellular function(s) in S. alaskensis.

6.4.3.2 Energy generation

The  and  subunits of the F0/F1 ATP-synthase appeared with increased intensity in chemostats at high growth rates (M152 and M001). The  and  subunits of the 173 Chapter 6

F0/F1 ATP-synthase also appeared with increased intensity during logarithmic growth

(Fegatella, 2002). This protein complex is the major ATP generating system in bacterial cells and is composed of membrane intrinsic and extrinsic sectors F0 and F1.

The  and  subunits are contained in the membrane extrinsic F1 sector. The  subunits are the non-catalytic nucleotide binding sites, which function cooperatively with the catalytic  subunits (Neidhardt et al., 1990). When ATP is abundant, ATP hydrolysis is favoured and the complex pumps protons across the membrane (Noji and Yoshida, 2000) thereby providing a proton gradient. When cellular ATP levels are low, the proton translocation reverses the rotation of the F0 sector resulting in

ATP synthesis.

The cellular levels of the two ATP synthase subunits are low during growth in chemostats at low growth rates relative to the increased levels at a high growth rate

(Table 6.6). At low growth rates the nutrient flux approaches complete starvation.

Through a range of adaptations, including high affinity transport systems (Schut et al., 1995) S. alaskensis has evolved an efficient capacity to convert substrates into biomass at low growth rates (Chapter 4). The lower abundance of the ATP synthase subunits indicates it is an important part of the adaptive response to low rates of growth in S. alaskensis, and is consistent with a reduced capacity of the cell to grow.

Moreover, it demonstrates that the down regulation does not impair the ability of the cell to maintain an efficient conversion of substrates. Since glucose serves as the main energy source in both glucose limited and batch cultures the decrease in the levels of ATP synthase  and  subunits during starvation and nutrient limitation is not surprising and the increase in the levels of these proteins is consistent with increased demand for energy generation. These results suggest that a tight coupling 174 Chapter 6

of these components with the availability of substrates or growth capacity is important in S. alaskensis.

The increased level of a putative electron transfer flavoprotein (M133) at low growth rate seems to indicate that electron transport may be important for the cells physiological state at low growth rate. It may be through a direct role in respiratory processes leading to energy generation, or indirectly by maintaining uptake or detoxification systems. Succinyl-coA-synthetase generates high energy phosphate intermediates in the TCA cycle. The increased level of this protein at high rates of growth is also consistent with higher energy and catabolism demands of the cell.

6.4.3.3 Sulphite reductase

The final match can be classified into other cellular processes. M117 matched the  subunit of a NADPH-sulphite reductase. The  subunit has a Mr of 64,000 Da in

Salmonella typhimurium whereas M117 was 37,000 Da. The protein predicted from thegenomeofS. aromaticivorans is also ~64,000 Da. It is unclear whether M117 has resulted from cleavage or whether it is encoded by a substantially shorter ORF.

Assimilatory sulphite reductases are highly conserved and are found in a wide range of bacteria, fungi and (Yamazaki et al., 1996). They generate sulphide for incorporation into sulphur containing amino acids and enzyme co-factors. The apparent increased expression at low growth rates may invoke a specific role for this protein nutrient acquisition and assimilation.

175 Chapter 7

7GENERAL DISCUSSION

176 Chapter 7

7.1 SUMMARY

With the overall aim of understanding the factors that control primary production and biogeochemical cycles in general there are four key objectives in marine microbiological research (i) to gain a better understanding of the distribution and abundance of different microbiological groups, (ii) to isolate ecologically relevant species for in-depth laboratory studies, in order to (iii) gain a better understanding of their physiologies and potential activities and (iv) use this knowledge to generate ecological hypotheses that can be tested with controlled experiments and physiological probes in order to understand their function in the ecosystem as a whole.

In this study a numerically abundant isolate from the North Pacific was characterised and found to be a genetically distinct strain of the model oligotrophic ultramicrobacterium Sphingopyxis alaskensis (Chapter 2). Methods and equipment

(small scale chemostats) were designed for the physical and molecular characterisation of marine isolates under conditions of nutrient limited growth

(Chapter 3). In conjunction with morphological measurements, biochemistry and stress resistance assays the equipment designed and tested in Chapter 3 was used to probe the physiology of S. alaskensis under conditions of nutrient limitation (Chapter

4). Model copiotrophic bacteria, V. angustum S14 and E. coli, were grown under the same conditions, and found to possess responses to nutrient limitation that were distinctly different to those of S. alaskensis. In an attempt to understand the molecular basis of the unique physiology of this organism complementary two dimensional gel electrophoresis (2DE) methods were developed along with a protein radiolabelling assay to observe changes in global gene expression (Chapter 5). High 177 Chapter 7

resolution 2DE expression maps were generated for S. alaskensis grown at high and low rates of growth in carbon limited chemostats and a number of proteins with altered expression profiles were identified by software analysis (Chapter 6). Finally a number of interesting proteins were further identified by mass spectrometry and cross species database matching, providing molecular insight into the basis of microbial oligotrophy. This work represents a solid scientific basis for further exploration of the molecular aspects of this model organism under controlled ecologically relevant conditions.

7.2 EXTINCTION DILUTION,

7.2.1 S. Alaskensis sp. strain AFO1

The extinction dilution method, as used to isolate strain AFO1, has proven to be successful for isolating abundant S. alaskensis ultramicrobacteria. A consistent finding has been the initial growth of cells only in liquid oligotrophic medium, with the ability to form colonies on solid media occurring after storage at 5°C for 6-12 months (Schut et al., 1997a) or storage and monthly reculturing at 15°C (Eguchi et al., 2001). The storage period enabled cells to form colonies not only on oligotrophic synthetic sea water medium (MPM) (Schut et al., 1997a), but also on rich synthetic sea water medium (ZoBell 2216E, Schut et al., 1997a, VNSS, Eguchi et al., 2001).

The initial lack of, but subsequent ability to grow on rich media indicates that the strains have facultative oligotrophic properties.

It is notable that even during the isolation of strain AFO1, only one of a total of 50 samples diluted to at least 10-5 supported growth on solid media, and the culture only 178 Chapter 7

developed this ability after serial passaging for a total of 12 months. During storage and culturing, a prerequisite for survival is therefore that appropriate nutrients have been included, and inhibitors excluded, and that sufficient time is taken to allow cells to grow.

7.2.2 Extinction dilution and lab domestication

A number of recent reports that utilise the extinction dilution method appear in the literature (Bernard et al., 2000, Eilers et al., 2000, Rappé et al., 2002, Connon and

Giovannoni, 2002) with mixed results. Very recently, Rappé et al. (2002) and

Connon and Giovannoni (2002) demonstrated the isolation of representatives of four major uncultured groups, including members of ubiquitous SAR11 lineage, with unamended seawater as media for extinction dilution culture. The SAR11 representative, Pelagibacter ubique HTCC1062, takes approximately 30 d to reach maximum cell numbers in sterile seawater medium, reaches a maximum of 3.5 x 106 cells per ml, does not grow on plates and its growth is inhibited by 0.001% proteose peptone. Strain AFO1 was able to form colonies on agar plates only after 12 months of regular subculture. S. alaskensis RB2256 required 6-12 months of incubation at

5˚C before an unknown mechanism enabled the cells to grow on agar plates.

All of these studies demonstrate that the critical factors for successful application of the extinction dilution method are long term culture, regular sub-culturing of dilution cultures and the use of low-nutrient unamended environmental water as media.

Taken together, each study suggests that the majority of organisms present in the environment are ‘alive’ but simply are unable to grow on the normally proffered

179 Chapter 7

substrates under the conditions provided. These observations point to nutrient status as a major factor affecting culturability. Further evidence in support of this comes from observations that vitamin B12 supplemented dilutions supported a 10-fold higher number of cells for samples taken from Toyama Bay, Japan (Mitsuru

Eguchi, unpublished), and Bruns et al., (2002) who demonstrate an increase in cultivation efficiency of bacteria from the Baltic sea by the inclusion of cAMP. The addition of cAMP has been shown to prevent substrate accelerated death in starved laboratory cultures (Calcot and Postgate, 1972) and the survival of S. alaskensis

RB2256 after exposure to stress may be affected by the type of nutrient used to limit growth (eg, malate and trehalose vs glucose) in chemostats and the type of agar medium used for recovery (This work and Amber Goodchild, Honours Thesis,

UNSW).

Despite recent success in isolating oligotrophic organisms from the marine environment it is important to consider what changes may occur during the lab domestication process and beyond. Throughout the course of this study care was taken to limit the possibility of mutations arising in the strain studied by regularly reviving the culture from a stock that was cryogenically preserved in 1992 in

Groningen, The Netherlands and by limiting the total number of generations in culture before returning to the original stock. It is indeed possible that mutants had arisen during the first year of the isolation process that enabled this strain to form colonies on agar plates and tolerate higher substrate concentrations. Such alterations could conceivably include mutations affecting the regulation of the levels of enzymes related to nutrient uptake, storage and osmoprotection or even so-called mutator alless (e.g. Notley-McRobb and Ferenci, 1999ab, Velicer et al., 1999) It is

180 Chapter 7

also possible that novel genes had been horizontally transferred from a or co-isolate.

A major part of this research (Chapters 4, 5 and 6) was motivated by a desire to understand the laboratory domestication process leading to the formation of colonies on plates (ie. a facultative oligotrophic state) from a state where no colonies were obtained (an obligately oligotrophic state). Initially this involved growing the strain at a low rate of growth in a nutrient limited chemostat in an attempt to reverse the switch to a facultative oligotroph (as assessed by the inability to form colonies on agar plates) and compare the differences in physiology. Although this approach was initially successful (F. Schut personal communication) we were unable to replicate the results. However, significant changes in the physiology of the organism grown at high and low rates of growth in chemostat were noted (Chapter 3), especially in comparison with the established copiotroph V. angustum S14. It was rationalised that low, nutrient limited rates of growth in chemostat were a better representation of the physiology of the original isolate in situ, in comparison with batch and high growth rates, and therefore this work focussed on documenting the differences in physiology between these different states and in direct contrast to well characterised copiotrophic bacteria where the exact experimental procedures were applied.

Current and future developments in large scale genome sequencing, metagenomics and genetic systems for environmental isolates are likely to provide the tools for investigating the role of specific genes and already some insights are emerging. For example, insertional inactivation of the gene for an amino acid transporter in

Synechococcus sp. WH 8102 (SYNW0828) relieves the toxicity of L-alanine, glycine

181 Chapter 7

and methionine in culture (B. Palenik, personal communication), suggesting that the lack of regulatory ability for that transporter can lead to a lethal accumulation of nutrients when the concentration of those nutrients is high. Future work of this type will provide a great benefit for the selection of suitable substrates, concentrations and overall approaches to isolating novel oligotrophic organisms. In addition to specific substrates a greater understanding of how slow nutrient limited growth affects the overall physiology of organisms is required, not only from the standpoint of isolating them from the environment but also to accurately assess their activities in situ.

7.3 THE IMPACT OF NUTRIENT LIMITATION ON THE PHYSIOLOGY OF MARINE BACTERIA

7.3.1 Distinct life strategies?

Through direct comparison of the physiology of model organisms in chemostat cultures the results presented in chapters 3 and 4 have implications for the distinctions between oligotrophs and copiotrophs. One critical observation is that V. angustum S14 continues to grow rather than differentiate into a dormant state when nutrients become limiting. Nutrient limitation in V. angustum S14 is accompanied by a decrease in cell size, a gradual increase in resistance to oxidative stress and an obvious decrease in cell yield. These results challenge the notion that environmental microcells are dormant copiotrophs that require a significant increase in cell size before initiation of cell division. Instead, a continuum of cell size, which is proportional to growth rate, is suggested and thus the distinction between copiotrophs and oligotrophs in situ on the basis of size and activity becomes less clear. One critical distinction remains, S. alaskensis does not suffer a decrease in growth efficiency (or maintenance energy demand) when nutrient limited, providing 182 Chapter 7

an obvious advantage in fitness for this organism over copiotrophic competitors in nutrient limited environments.

The fact that copiotrophs may be able to grow under extremely nutrient limited conditions in the absence of competition (Chapter 3) does not suggest that they are active members of the pelagic marine environment. In fact it seems likely that they would experience true starvation when in competition for limited nutrients with the indigenous bacteria that are truly adapted to growth in such extreme conditions. It seems likely that all marine bacteria are capable of surviving and growing with low concentrations of nutrients in the absence of competition however in the pelagic marine environment copiotrophs may be outcompeted by the truly low nutrient adapted bacteria. It seems plausible that for Vibrio at least there is some overlap between the responses to starvation and nutrient limitation.

In this regard it is interesting that even though S. alaskensis is physiologically suited to growth in oligotrophic conditions, it is genetically geared to respond to carbon starvation (Fegatella and Cavicchioli, 2000). It is likely, in terms of total available carbon and gradients of nutrients present in microzones, that starvation in the marine environment is of potential significance to all bacterioplankton. Clearly however, S. alaskensis has evolved a genotype that enabled it to be a numerically dominant bacterium at the time of its isolation, in contrast to the presence of significantly lower number (<1%) of copiotrophic bacteria (Button et al., 1993). It is therefore not surprising that, in contrast to copiotrophic bacteria which are likely to survive in oligotrophic waters by attachment to nutrient rich particles or by producing stress resistant, resting stage cells when detatched, S. alaskensis is likely to have developed

183 Chapter 7

optimal stress resistant mechanisms for survival during slow growth. It is important to consider that a coordinated and specific glucose starvation response has been described in S. alaskensis (Fegatella et al., 2000) which supports the classification of this organism as a facultative oligotroph. An explanation that integrates the results presented in this thesis with the work on starvation responses in S. alaskensis may be that V. angustum S14 and S. alaskensis display similar responses to starvation but at significantly different thresholds of nutrient deprivation.

7.3.2 Comparison of S. alaskensis with Peligibacter ubique

Culture-independent methods have established that Pelagibacter ubique , formerly identified as the SAR11 cluster, as probably the most abundant single species on earth (Rappé et al., 2002). Since the first identification of SAR11 16S rDNA sequences in the Sargasso Sea (Giovannoni et al., 1990) to the eventual isolation of the organism in 2002 by dilution culture, Pelagibacter ubique signature sequences have been found in clone libraries and FISH samples prepared from a variety of globally distributed sites and, at the time of isolation, accounted for 26% of all rRNA gene sequences identified in sea water (Rappé et al., 2002). Pelagibacter cells have dimensions of 0.37-0.89 m length and a diameter of only 0.12-0.20 m. With a volume of approximately 0.1 m3 these cells represent the smallest free-living cells isolated.

P. ubique maintains an intolerance of high concentrations of organic nutrients in liquid culture and is incapable of forming colonies on plates. The isolated strain displays a maximun rate of growth of 0.58 d-1 which is 10-fold lower than S. alaskensis (maximum rate of growth > 5.0 d-1, Chapter 3). This comparison clearly

184 Chapter 7

defines S. alaskensis as a facultative oligotroph versus P. ubique which approximates more closely to an obligate oligotroph. Unfortunately the slow rate of growth, lack of colonies on plates and difficulties in generating biomass severely limit the ability to probe the molecular physiology of this organism by way of physiological and genetic techniques.

P.ubique and S. alaskensis are both members of the ubiquitous -proteobacteria.

Despite the fact that both organisms encode a single copy of the rRNA operon, the genome size of S. alaskensis (3.345 Mbp plus a of 28.5 kb, www.jgi.doe.gov) is 2.5 times larger than that of P. ubique (1.309 Mbp, Giovannoni et al., 2005). The larger gene complement and higher maximum rates of growth may allow a single population of S. alaskensis to adapt to a wider variety of changing conditions and therefore persist in relatively high abundances in a water body throughout seasonal changes or perturbations, while the smaller gene complement of

P. ubique and lower maximum rate of growth may result in this organism being restricted to a narrow set of conditions. Even so, P. ubique sequences are globally distributed and found in high abundance in a variety of ocean systems which suggests that the small genome size and minimal growth requirements of the organism offer a considerable competitive advantage over a wide range of environmental conditions. An alternative explanation for the apparent of this single ‘species’ may be the existence of genetically distinct ‘’ of P. ubique (Morris et al., 2005), analogous to Prochlorococcus ecotypes (Moore et al.,

1998), whose gene complements differ through processes of lateral gene transfer and gene loss (Coleman et al., 2006).

185 Chapter 7

S. alaskensis strains have also been isolated from globally distant sites (Chapter 3).

Strains AFO1 and RB2256 display an amazing degree of genome sequence identity as observed by a DNA-DNA hybridisation value of 84%, a value that suggests an average nucleotide identity (ANI) of > 95% at the genome sequence level

(Konstantinidis and Tiedje, 2005). In contrast, Prochlorococcus ecotypes display

ANI% values on the range of 78.4-79.5%, which is incongruent with the > 97% identity between these strains at the 16S rDNA level. Taken together with large numbers of unique genes found in the genomes of each ecotype (Coleman et al.,

2006, Rocap et al., 2003, Moore et al., 2005) these observations suggest that such ecotypes are evolutionarily distant, genetically distinct and that differences in gene complement in each ecotype enable the ‘species’ to occupy a broader range of environments. In the case of Prochlorococcus it appears that the nearly minimal genome is a result of fixation of each ecotype at a particular set of environmental conditions, or niche. On the other hand, the larger gene complement in S. alaskensis may have enabled this single species to adapt to a wider range of physiological conditions which is evident from the nearly identical strains AFO1 and RB2256 which colonise geographically distinct sites in the temperate waters off Japan

(Chapter 3) and Resurrection Bay, Alaska.

7.4 THE MECHANISMS OF STRESS RESISTANCE IN S. ALASKENSIS

In S. alaskensis nutrient limitation results in a smaller cell size, and a slower growth rate and notably a dramatic increase in the observed resistance to oxidative stress.

Cellular levels of the molecular chaperone GroEL are also elevated at low rates of growth. Taken together with the high growth yield at low rates of growth these results imply that the cell has selected a strategy of protection, rather than 186 Chapter 7

replacement, of cellular components that are susceptible to damage during slow growth or starvation. These observations suggest that the constitutive expression of, what would normally be perceived as, metabolically expensive general stress protective mechanisms, in copiotrophs, is more economical in S. alaskensis than replacing damaged components by new synthesis and growth. Even so, some cellular proteins are observed to have apparently high synthesis rates but low overall abundance, notably the S1-S5 cluster, during nutrient limited growth (Chapters 5 and

6) and starvation (Fegatella and Cavicchioli, 2000) suggesting that these proteins are susceptible to damage and are rapidly turned-over. It is tempting to speculate that S1-

S5 is an important but sensitive component of the transport machinery that is rapidly turned-over despite nutrient deprivation and slow growth.

The ability of S. alaskensis to immediately respond to the addition of nutrients is another obvious competitive advantage (Fegatella et al., 1998, Eguchi et al., 1996).

Since the uptake of nutrients through specific transporters is energy dependent we would expect a considerable amount of resources diverted to maintaining an energised membrane, possibly through access to carbon and/or energy stores.

Alternatively, S. alaskensis may rely on the immediate conversion of captured substrates to energy rather than conversion to stores by default. The benefit here would be a faster response through the reduction of regulated responses and regulatory mechanisms. This indeed is reflected by constitutive uptake of a broad range of substrates and the apparent lack of catabolite repression in this organism.

One obvious drawback however, is an unpredictable and variable flow through the electron transport chain coupled to the unpredictable supply of nutrient. Therefore there would be an increased likelyhood of generating reactive oxygen species when

187 Chapter 7

the transport chain is over-oxidised during nutrient pulses. The induction of protective mechanisms at low rates of growth could be a mechanism for coping with reactive oxygen species generated in this scenario.

The potential sources of reactive oxygen species, and indeed the main targets of damage, have not been determined in S. alaskensis or any other environmental isolate. The experiments done here highlight the common importance of stress resistance mechanisms at low rates of growth for oligotrophs and copiotrophs (E. coli and V. angustum S14, Chapter 3 and Chapter 4), indicating that more detailed experiments are likely to provide insight into the factors that impact on the fitness of organisms in nutrient limited environments. For S. alaskensis the use of various hydroperoxides with specific targets, such as linoleic acid hydroperoxide which targets membrane components (Evans et al., 1998), and immunodetection and identification of damaged proteins (e.g. Shanmuganathan et al., 2004) could offer useful approaches for more detailed studies to define the critical cellular targets of reactive oxygen species.

7.4.1 Carotenoids

The data presented at the end of Chapter 4 forms the basis of investigations into the link between the cellular pigments of S. alaskensis and a range of interesting physiological features which include, the ability to withstand high levels of oxidative stress and UV radiation in all growth conditions and broad spectrum resistance to antibiotics. There is some evidence that carotenoids can protect , and microbial cells from photooxidative damage by scavenging reactive oxygen species

(Tuveson and Sandmann, 1993), Although the protection of cellular componets, 188 Chapter 7

such as DNA, from damage by direct absorbtion of UV wavelengths has not been experimentally determined, and is probably unlikely since the absorption maxima of carotenoids are usually found in the wavelength range above 400 nm.

Although there was no correlation between the overall abundance of carotenoids and the observed bimodal response to hydrogen peroxide, the induction of a mechanism of chemical modification, such as glycosylation, might explain how, below a certain threshold, the level of nutrient limitation acts like a switch resulting in the two distinct levels of resistance to hydrogen peroxide. More detailed investigations are needed in order to substantiate the possible roles of carotenoids in protection against oxidative stress. Evidence in support of this hypothesis may be obtained from ongoing work focussing on the examination of chemical modifications and isomerisation of carotenoid components purified by HPLC. The presence of the compound with a strong absorbance maximum at 230 nm in intact cells and excreted to the growth medium is intriguing and begs further investigation. Due to the small size and large surface area to volume ratio and the requirements for nutrient scavenging and electron transport machinery, the membranes of S. alaskensis are likely to be rich in protein. Thus carotenoids may play an important structural role in ensuring membrane integrity (Britton, 1995) and future work should be directed towards understanding their function in S. alaskensis and other marine isolates.

7.5 GENE EXPRESSION DURING NUTRIENT LIMITED GROWTH AND STARVATION

A summary of protein spots with altered expression levels during slow nutrient limited growth or in response to glucose starvation (Fegatella and Cavicchioli, 2000)

189 Chapter 7

is presented on the combined 2DE protein reference maps for S. alaskensis in Figure

7.1. These reference maps form the basis for continued expoloration of the unique physiology of this model marine oligotroph. The relatively low numbers of proteins with differential spot intensities on radiolabelled and silver stained gels observed at high and low rates of growth in chemostats indicates that nutrient limitation does not induce major changes in gene expression in S. alaskensis. In direct contrast, glucose starvation elicits global changes in more than 20 % of analysed spots on 2DE images

190 FIGURE 7.1. Combined 2DE protein reference maps for S. alaskensis RB2256. The predicted molecular mass and pI values of specific spots were assigned by Z3 software (Compugen). Highlighted spots correspond to those with altered expression levels or those identified by cross species database matching (chapter 6 and Fegatella PhD Thesis, UNSW). 191 Chapter 7

(Figure 7.1, Fegatella and Cavicchioli, 2001). However, the physiology of glucose starved cells indicate that a majority of the observed changes are likely to be due to the down-regulation of proteins involved in translation and transcription and other cellular processes rather than induction of starvation-induced proteins.

Although it is disappointing that very few starvation or low nutrient-induced proteins have been identified, S. alaskensis proteomics projects support the observed lack of phenotypic responses to a wide range of nutrient concentrations. These results support the hypothesis that systems for high affinity uptake of diverse nutrients and stress resistance mechanisms are constitutively expressed in S. alaskensis and that the protection and maintenance of existing cellular machinery, aided by the expression of molecular chaperones, is energetically or competitively more favourable than turn-over of damaged components. Constitutive expression of the genome can also create problems for an organism by removing an important level of regulation, e.g. a cell may not be able to regulate the flux through biochemical pathways. Given the relatively high degree of post-translational modifications observed on 2DE profiles of S. alaskensis (Chapters, 5, 6 and Fegatella and

Cavicchioli, 2000) it is entirely possible that S. alaskensis maintains homeostasis by post-tranlational, rather than transcriptional regulation. If the emphasis of regulation has indeed been shifted from transcription to translation this would be evident by a relatively low number of two-component signal transduction systems, transcriptional regulators and alternate sigma factors in the genome sequence of the organism. It would also highlight the advantages of physiological and proteomics approaches over transcriptomics (i.e. microarrays) for dissecting the molecular biology of this model oligotroph. The reproducibility of 2DE protein profiles in S. alaskensis show

192 Chapter 7

promise for the use of this method for assessing the physiological impact of future targeted gene deletions in chemostats or batch cultures (discussed below) in conjunctin with the 2DE reference maps presented here (Figure 7.1).

In summary this work supports the classification of S. alaskensis as a facultative oligotroph at the molecular level and clearly highlights some interesting distinctions between the lifestyles of oligotrophic, mesotrophic and copiotrophic heterotrophic bacteria in response to nutrient restriction.

7.6 FUTURE PROSPECTS

7.6.1 Molecular genetics

A lot of progress has been made in the development of suitable tools for the genetic manipulation of S. alaskensis. A genetics system consisting of RSF1010 -based replicating vectors and tn5 derivative transposons delivered by tri- parental conjugation has been developed for S. alaskensis (Cavicchioli, unpublished).

S. alaskensis remains resistant to high concentrations of a wide variety of antibiotics, including kanamycin, gentamycin and streptomycin. At present chloramphenicol and ampicillin remain the only selectable markers, although the minimal inhibitory concentration for ampicillin is relatively high (>1.0 mg ml-1). Nevertheless, transposons with chloramphenicol resistance cassettes are integrated into the genome and stably maintained (K. Kandelbinder, Honours Thesis, UNSW), paving the way for mutagenesis by random integration followed by suitable plate or liquid screens for interesting phenotypes. The replicative plasmids pSEB24 (Yamazaki et al., 1996) and pCJS10 (Charles Svenson, unpublished) are also stably maintained and transferred with high frequency to S. alaskensis though conjugation. A fluorescence-

193 Chapter 7

activated cell sorting FACS-optimised green fluorescent protein (Cormack et al.,

1996), expressed from a constitutive promoter on a plasmid vector (pCJS10 and derivatives), has also been used as a selectable marker for S. alaskensis in conjunction with flow cytometry. Although this has not yet been demonstrated, the success with transposons and replicative plasmids suggests that targeted mutations may also be possible through homologous recombination with suitable constructs containing the chloramphenicol cassette delivered on a non-replicating plasmid. In combination with the proteomics methods developed here and the completed genome sequence such genetics tools provide important resources for unravelling the molecular basis the unique physiology of S. alaskensis.

7.6.2 Cross species identification of proteins: environmental proteomics

In the absence of genome sequence data for S. alaskensis this study demonstrates the success of cross species database matching for identifying relatively conserved proteins from 2DE gels using mass spectrometry. At present the S. alaskensis genome sequence has been completed (www.jgi.doe.gov), making the cross species approach obsolete for laboratory studies of this organism. Even so, the cross species matching approach represents a significant step-forward in that it provides a tool with which we can begin to contemplate proteomics on an environmental scale. In other words, cross-species matching should enable the identification of environmentally expressed and abundant proteins from environmental samples, thereby revealing a snap-shot of the biological processes underway and providing an assessment of the physiological state of a community at a particular time and place.

194 Chapter 7

The reference genomic data that is required for an ‘environmental proteomics’ approach will be provided by on-going sequencing initiatives, such as the Gordon and Betty Moore Foundation Marine Microbiology Initiative (10 X coverage of 150 marine microbial genomes, www.moore.org), sequencing of environmentally important and phylogenetically diverse genomes at the Joint Genome Institute and

Genoscope as well as past and future environmental shotgun (Venter et al., 2004) and BAC library sequencing. Given the substantial amount of environmental sequence already generated from the Venter et al. (2004) Sargasso Sea project and the number of marine genomes that are either completed (eg. Palenik et al., 2003,

Dufresne et al., 2003) or scheduled (see www.moore.org, www.gold.org), the marine environment is an obvious candidate for an environmental proteomics approach. Any attempt at environmental proteomics would benefit enormously from recent advances in protein separation whereby mixtures of proteins are separated by multi- dimensional liquid chromatography prior to LC-ESI-MS/MS. Clearly, more in- depth physiological analysis of environmentally important isolates will be needed to link the roles of physiologically relevant protein markers and the stress or starvation status of environmental populations.

7.7 CONCLUDING REMARKS

Recent isolations of previously uncultured, yet abundant, members of ocean microbial communities, such as Pelagibacter ubique, has provided a range of genetically diverse important oligotrophic isolates for laboratory studies. In-depth physiological and genetic analyses of such isolates, including S. alaskensis, under environmentally relevant conditions is essential for testing and generating

195 Chapter 7

environmental hypotheses and understanding the fundamental processes that shape ocean food webs and global biogeochemical cycles.

196 8 REFERENCES

Alldredge,A.L.,andM.W.Silver.1988. Characteristics, dynamics and significance of marine snow. Progress Oceanography 20:41-82.

Arana, I., A. Muela, J. Iriberri, L. Egea, and I. Barcina. 1992. Role of hydrogen peroxide in loss of culturability in Escherichia coli in a . Applied and Environmental Microbiology 58:3903-3907.

Armstrong, G. 1997. Genetics of eubacterial carotenoid biosynthesis: a colourful tale. Ann. Rev. Microbiol. 51:629-659.

Azam, F. 1998. Microbial control of ocean carbon flux: the plot thickens. Nature 280:694-696.

Azam, F., and J. W. Ammerman. 1984. Cycling of organic matter by bacterioplankton in pelagic : microenvironmental consideration, p. 345-360. In M. J. R. Fasham (ed.), Flows of energy and materials in marine ecosystems: theory and practice. Plenum Publishing, New York.

Balkwill, D.L., G. R. Drake,R. H. Reeves,J. K. Fredrickson, D. C. White, D. B. Ringelberg, D. P. Chandler, M. F. Romine, D. W. Kennedy, and C. M. Spadoni. 1997 Taxonomic study of aromatic-degrading bacteria from deep- terrestrial- subsurface sediments and description of Sphingomonas aromaticivorans sp. nov., Sphingomonas subterranea sp. nov., and Sphingomonas stygia sp. nov. Int J Syst Bacteriol 47: 191-201.

Baraniecki ,C. A., Aislabie, J., Foght J.M. 2002. Characterization of Sphingomonas sp. Ant 17, an aromatic hydrocarbon-degrading bacterium isolated from Antarctic soil. Microbial Ecology. 43:44-54.

Beers, R.F. and I.W. Sizer. 1952. A spectrophotometric method for measuring the breakdown of hydrogen peroxide by catalase. Journal of Biological Chemistry 195:133-140.

Berg, Ø., Børsheim, K. Y., G. Bratbak, and M. Heldal. 1989. High abundance of found in aquatic environments. Nature 340:467-468.

Bernard, L., C. Courties, P. Servais, M. Troussellier, M. Petit, and P. Lebaron. 2000. Relationships among bacterial cell size, productivity, and genetic diversity in aquatic environments using cell sorting and flow cytometry. Microbiolgy and Ecology. 40:148-158.

Bernard, L., H. Schafer, F. Joux, C. Courties, G. Muyzer, and P. Lebaron. 2000. Genetic diversity of total, active and culturable marine bacteria in coastal seawater. Aquatic Microbial Ecology 23:1-11.

197 Blackburn, N., T. Fenchel, and J. Mitchell. 1998. Microscale nutrient patches in planktonic habitats shown by chemotactic bacteria. Science 282:2254-2255.

Bligh, E.G., and W.J. Dyer. 1959. A rapid method for total lipid extraction and purification. Canadian journal of Biochemistry and Physiology 37:911-917.

Bowman, J. P., S. A. McCammon, J. L. Brown, and T. A. McMeekin. 1998. Glaciecola punicea gen. nov., sp. nov.: psychrophilic bacteria from Antarctic sea-ice habitats. International Journal of Systematic Bacteriology 48:1213- 1222.

Bridges, B. A., and A. Timms. 1998. Effect of endogenous carotenoids and defective RpoS sigma factor on spontaneous mutation under starvation conditions in Escherichia coli - evidence for the possible involvement of singlet oxygen. Mut. Res. 403:21-28.

Britschgi, T. B., and S. J. Giovannoni. 1991. Phylogenetic analysis of a natural marine bacterioplankton population by rRNA gene cloning and sequencing. Applied and Environmental Microbiology 57:1707-1713.

Britton, G. 1995. The structure and properties of carotenoids in relation to function FASEB J. 9:1551-1558.

Britton, G. 1995b. UV/Visible spectroscopy. In: G. Britton, S. Liaaen-Jensen and H. Pfander (Eds). Carotenoids, Vol1b. Birkhauser, Basel. pp 13-62.

Bruns, A., H. Cypionka, and J. Overmann. 2002. Cyclic AMP and acyl homoserine lactones increase the cultivation efficiency of heterotrophic bacteria from the central Baltic Sea. Applied and Environmental Microbiology 68:3978-3987

Büttner K., J. Bernhardt, C. Scharf, R. Schmid, U. Mader, C. Eymann, H. Antelmann, A. Volker, U. Volker and M. Hecker 2001. A comprehensive two-dimensional map of cytosolic proteins of Bacillus subtilis. Electrophoresis. 22:2908-2935.

Button, D. K. 1991. Biochemical basis for whole-cell uptake kinetics: specific affinity, oligotrophic capacity and meaning of the Michaelis constant. Applied and Environmental Microbiology 57:2033-2038.

Button, D. K., and B. R. robertson. 1989. Kinetics of bacterial processes in natural aquatic systems based on biomass as determined by high resolution flow cytometry. Cytometry 10:558-563.

Button, D. K., and B. R. Robertson. 2000. Effect of nutrient kinetics and cytoarchitecture on bacterioplankter size. Limnology and Oceanography 45:499-505.

Button, D. K., and B. R. Robertson. 2001. Determination of DNA content of aquatic bacteria by flow cytometry. Applied and Environmental Microbiology 67:1636-1645. 198 Button, D. K., B. R. Robertson, P. W. Lepp, and T. M. Schmidt. 1998. A small, dilute-cytoplasm, high-affinity, novel bacterium isolated by extinction culture and having kinetic constants compatible with growth at ambient concentrations of dissolved nutrients in seawater. Applied and Environmental Microbiology 64:4467-4476.

Button,D.K.,F.Schut,P.Quang,R.Martin,andB.R.Robertson.1993. Viability and isolation of marine bacteria by dilution culture : theory, procedures and initial results. Applied and Environmental Microbiology 59:881-891.

Campbell,L.,H.A.Nolla,andD.Vaulot.1994. Photosynthetic picoplankton community structure in the subtropical North Pacific Ocean. Limnology and Oceanography 34:954- 961.

Carlson, C. A., and H. W. Ducklow. 1996. Growth of bacterioplankton and consumption of dissolved organic carbon in the Sargasso Sea. Aquatic Microbial Ecology 10:69-85.

Carlson, C. A., H. W. Ducklow, and T. D. Sleeter. 1996. Stocks and dynamics of bacterioplankton in the northwestern Sargasso Sea. Deep-Sea Research II 43.

Caron, D. A., E. R. Peele, E. L. Lim, and M. R. Dennett. 1999. Picoplankton and nanoplankton and their trophic coupling in surface waters of the Sargasso Sea south of Bermuda. Limnology and Oceanography 44:259-272.

Cash, P. 1998. Characterisation of bacterial proteomes by two-dimensional electrophoresis. Analytica Chimica Acta 372:121-145.

Cashel, M., D. R. Gentry, V. J. Hernandez, and D. Vinella. 1996. The stringent response, p. 1456-1496. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella : cellular and molecular biology, 2nd ed, vol. 1. American Society for Microbiology, Washington, D. C.

Cavender-Bares, K.K., Karl, D.M., and Chisholm, S.W. 2001 Nutrient gradients in the western North Atlantic Ocean: Relationship to microbial community structure and comparison to patterns in the Pacific Ocean. Deep-Sea Research Part I-Oceanographic Research Papers 48: 2373-2395.

Cavicchioli, R., F. Fegatella, M. Ostrowski, M. Eguchi, and J. Gottschal. 1999. Sphingomonads from marine environments. Journal of Industrial Microbiology and Biotechnology 23:268-272.

Chatellier, J., F. Hill, P. A. Lund, and A. R. Fersht. 1998. In vitro activities of GroEL minichaperones. Proceddings of the National Academy of Science USA 95:9861-9866.

199 Chesbro, W. 1988. The domains of slow bacterial growth. Canadian Journal of Microbiology 34:427.

Chesbro, W. R., T. Evans, and R. Eifert. 1979. Very slow growth of Escherichia coli. Journal of Bacteriology. 139:625-638.

Chesbro, W. R., T. Evans, and R. Eifert. 1990. When nutrient limitation places bacteria in the domains of slow growth: metabolic, morphologic and cell cycle behaviour. FEMS Microbiol. Ecol. 74:103.

Chin,W-C,M.V.OrellanaandP.Verdugo.1998 Spontaneous assembly of marine dissolved organic matter into polymer gels. Nature 391: 568 - 572.

Cho, B. C., and F. Azam. 1988. Major role of bacteria in biogeochemical fluxes in the oceans interior. Nature 322:441-443.

Cho, B. C., and F. Azam. 1990. Biogeochemical significance of bacterial biomass in the ocean's euphotic zone. Marine Ecology Progress Series 63:253-259.

Cho, J.-C., and Giovannoni, J. 2004 Cultivation and Growth Characteristics of a Diverse Group of Oligotrophic Marine Gammaproteobacteria. Applied and Environmental Microbiology 70: 432-430. Christner, B.C., E. Mosley-Thompson, L.G. Thompson and J.N. Reeve. 2001. Isolation of bacteria and 16S rDNAs from accretion ice. Environmental Microbiology. 3(9):570-577.

Church, M. J., D. A. Hutchins, and H. W. Ducklow. 2000. Limitation of bacterial growth by dissolved organic matter and iron in the Southern Ocean. Applied and Environmental Microbiology 66:455-466.

Coffin, R. B. 1989. Bacterial uptake of dissolved free and combined amino acids in estuarine waters. Limnology and Oceanography 34:531-542.

Cole, J. J., S. Findlay, and M. L. Pace. 1988. Bacterial production in fresh and saltwater ecosystems: A cross-system overview. Marine Ecology Progress Series 43:1-10.

Coleman,M.L.,M.B.Sullivan,A.C.Martiny,C.Steglich,K.Barry,E.F. DelongandS.W.Chisholm.2006. Genomic islands and the ecology and evolution of Prochlorococcus. Science 311: 1768-70.

Connon, S. A., and S. J. Giovannoni. 2002. High-throughput methods for culturing microorganisms in very-low-nutrient media yield diverse new marine isolates. Applied and Environmental Microbiology 68:3878-85.

Cooper, W. J., and R. G. Zika. 1983. Photochemical formation of hydrogen peroxide in surface and ground waters exposed to sunlight. Science 220:711- 71.

200 Cordwell,S.J.,D.J.Basseal,andI.Humphery-Smith.1997. Proteome analysis of Spiroplasma melliferum (A56) and protein characterisation across species boundaries. Electrophoresis 18:1335-1346.

Cordwell,S.J.,V.C.Wasinger,A.Cerpa-Poljak,M.W.Duncan,andI. Humphery-Smith. 1997. Conserved motifs as the basis for recognition of homologous proteins across species boundaries using peptide-mass fingerprinting. J Mass Spectrom 32:370-8.

Cordwell,S.J.,M.R.Wilkins,A.Cerpa-Poljak,A.A.Gooley,M.Duncan,K.L. Williams, and I. Humphrey-Smith. 1995. Cross-species identification separated by two-dimensional gel electrophoresis using matrix assisted laser desorption ionization/time-of-flight mass spectrometry and amino acid composition. Electrophoresis 16:438-443.

Cormack, B. P., R. H. Valdivia, and S. Falkow. 1996. FACS-optimised mutants of the green fluorescent protein (GFP). Gene 173:33-38.

Corthals, G., S. P. Gygi, R. Aebersold, and S. D. Patterson. 1999. Identification of protein by mass spectrometry, p. 197-231. In T. Rabilloud (Ed.), Proteome research: 2D gel electrophoresis and detection methods. Springer, New York.

Corthals, G., M. P. Molloy, B. R. Herbert, K. L. Williams, and A. A. Gooley. 1997. Prefractionation of protein samples prior to two-dimensional electrophoresis. Electrophoresis 18:317-323.

Corthals, G. L., V. C. Wasinger, D. F. Hochstrasser, and J. Sanchez. 2000. The dynamic range of protein expression: a challenge for proteomic research. Electrophoresis 21:1104-1115.

Cotner, J. B., A. J. W., E. R. Peele, and E. Bentzen. 1997. Phosphorus-limited bacterioplankton growth in the Sargasso Sea. Aquatic Microbial Ecology 13:141-149.

Cottrell, M. T., and D. L. Kirchman. 1992. Natural assemblages of marine proteobacteria and members of the Cytophaga-Flavobacter cluster consuming low- and high-molecular-weight dissolved organic matter. Applied and Environmental Microbiology 66:1692-1697.

Cottrell, M. T., and D. L. Kirchman. 2000. Community composition of marine bacterioplankton determined by 16S rRNA gene clone libraries and fluorescence in situ hybridization. Applied and Environmental Microbiology 66:5116-5122.

Cottrell, M. T., D. N. Wood, L. Y. Yu, and D. L. Kirchman. 1995. Selected chitinase genes in cultured and uncultured marine bacteria in the alpha- and gamma-subclasses of the proteobacteria. Applied and Environmental Microbiology 66:1195-1201.

201 Courchesne, P. L., and S. D. Patterson. 1997. Identification of proteins by matrix- assisted laser desorption/ionization mass spectrometry using peptide and fragment ion masses. Methods in Molecular Biology 112:487-512.

Davis,B.D.,S.M.LugerandP.C.Tai.1986. Role of ribosome degradation in the death of starved Escherichia coli cells. Journal of Bacteriology. 166:439-445.

Davison, A. D., M. R. Gillings, D. R. Jardine, P. Karuso, A. S. Nouwens, J. J. French, D. A. Veal and N. Altavilla. 1999. Sphingomonas paucimobilis BPSI-3 mutant AN2 produces a red catabolite during biphenyl degradation. Journal of Industrial Microbiology and Biotechnology 23:314-319. del Giorgio, P. A. and J. J. Cole. 2000. Bacterial energetics and growth efficiency. 289-325 pp, in D. Kirchman (ed.), Marine Microbial Ecology. Plenum Press. del Giorgio, P. A., and G. Scarborough. 1995. Increase in the proportion of metabolically active bacteria along gradients of enrichment freshwater and marine plankton: implications for estimates of bacterial growth and production rates. Journal of Plankton Research 17:1905-1924.

DeLong, E. F., D. G. Franks, and A. L. Alldredge. 1993. Phylogenetic diversity of aggregate-attached vs. free-living marine bacterial assemblages. Limnology and Oceanography 38:924-934.

DeLong, E. F., L. T. Taylor, T. L. Marsh, and C. M. Preston. 1999. Visualization and enumeration of marine planktonic archaea and bacteria by using polyribonucleotide probes and fluorescent in situ hybridization. Applied and Environmental Microbiology 65:5554-5563.

Douglas,D.J.,J.A.Novitsky,andR.O.Fournier.1987. Microautoradiography- based enumeration of bacteria with estimates of thymidine-specific growth and production rates. Marine Ecology Progress Series 36:91-99.

Ducklow, H. W. 1999. The bacterial component of the oceanic euphotic zone. FEMS Microbial Ecology 30:1-10.

Ducklow, H. W., D. L. Kirchman, H. L. Quinby, C. A. Carlson, and H. G. Dam. 1993. Stocks and dynamics of bacterioplankton carbon during the spring bloom in the eastern North Atlantic Ocean. Deep-Sea Research II 40:245-263.

Ducklow, H. W., H. L. Quinby, and C. A. Carlson. 1995. Bacterioplankton dynamics in the equatorial Pacific during 1992 El Nino. Deep-Sea Research II 42:621- 638.

Dunn, M. J. 1999. Electroblotting of proteins from 2-D polyacrylamide gels. In A. J. Link (ed.), Methods in Molecular Biology, vol. 112: 2-D Proteome Analysis Protocols. Humana Press Inc., Totawa, N.J.

202 Eberl, L., M. Givskov, C. Sternberg, S. Moller, G. Christiansen, and S. Molin. 1996. Physiological Responses of Pseudomonas putida KT2442 to Phosphate Starvation. Microbiology 142:155-163.

Eguchi, M. 1999. The nonculturable state of marine bacteria. In C. R. Bell, M. Brylinski, and P. Johnson-Green (ed.), Microbial biosystems: new frontiers. Atlantic Canada Society for Microbial Ecology, Halifax, Canada.

Eguchi, M., and A. Kawai. 1992. Distribution of oligotrophic and eutrophic bacteria in fish culturing inland bays. Nippon Suisan Gakkaishi 58: 119-125.

Eguchi, M., and Y. Ishida. 1990. Oligotrophic properties of heterotrophic bacteria and in situ heterotrophic activity in pelagic seawaters. FEMS Microbiology Ecology 73:23-30.

Eguchi, M., T. Nishikawa, K. MacDonald, R. Cavicchioli, J. C. Gottschal, and S. Kjelleberg. 1996. Responses to stress and nutrient availability by the marine ultramicrobacterium Sphingomonas sp. strain RB2256. Applied and Environmental Microbiology 62:1287-1294.

Eilers, H., J. Pernthaler, F. O. Glockner, and R. Amann. 2000. Culturability and In situ abundance of pelagic bacteria from the North Sea. Applied and Environmental Microbiology 66:3044-51.

Eng, J., K., A. McCormack, L. , and J. R. Yates, III. 1994. An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. Journal of American Society for Mass Spectrometry 5:976- 989.

Evans M. V., H. E. Turton, C. M. Grant, and I. W. Dawes. 1998. Toxicity of Linoleic Acid Hydroperoxide to Saccharomyces cerevisiae: Involvement of a Respiration-Related Process for Maximal Sensitivity and Adaptive Response. Journal of Bacteriology 180: 483-490.

Farr, S. B., and T. Kogoma. 1991. Oxidative stress responses in Escherichia coli and Salmonella typhimurium Microbiology Reviews 55.

Fegatella, F. 2002. Physiological adaptation to carbon starvation in a marine oligotrophic ultramicrobacterium Sphingomonas alaskensis RB2256. PhD Thesis. University of New South Wales.

Fegatella, F., and R. Cavicchioli. 2000. Physiological responses to starvation in the marine oligotrophic ultramicrobacterium Sphingomonas sp. strain RB2256. Applied and Environmental Microbiology 66:2037-2044.

Fegatella, F., J. Lim, S. Kjelleberg, and R. Cavicchioli. 1998 Implications of rRNA operon copy number and ribosome content in the marine oligotrophic ultramicrobacterium Sphingomonas sp. strain RB2256. Applied and Environmental Microbiology 64:4433-4438.

203 Fegatella, F., M. Ostrowski, and R. Cavicchioli. 1999. An assessment of protein profiles from the marine oligotrophic ultramicrobacterium, Sphingomonas sp. strain RB2256. Electrophoresis 20:2094-2098.

Felsenstein, J. 1985. Confidence limits on phylogenies: an approach using the bootstrap. Evolution 39:166-170.

Ferenci, T. 1999. Regulation by nutrient limitation. Current Opinion in Microbiology 2:208-213.

Ferguson, R. L., E. N. Buckley, and A. V. Palumbo. 1984. Response of marine bacterioplankton to differential filtration and confinement. Applied and Environmental Microbiology 47:49-55.

Ferguson, G.P., Creighton, R.I., Nikolaev, Y., and Booth, I.R. 1998. Importance of RpoS and Dps in survival of exposure of both exponential- and stationary- phase Escherichia coli cells to the electrophile N-ethylmaleimide. Journal of Bacteriology 180:1030-1036.

Flärdh, K., T. Axberg, N. H. Albertson, and S. Kjelleberg. 1994. Stringent control during carbon starvation of marine Vibrio sp. strain S14: molecular cloning, nucleotide sequence and deletion of the relA gene. Journal of Bacteriology 176:5949-5957.

Fong,N.J.C.,M.L.Burgess,K.D.Barrow,andD.R.Glenn.2001. Carotenoid accumulation in the psychrotrophic bacterium Arthrobacter agilis in response to thermal and salt stress. Applied Microbiology and Biotechnology 56:750- 756.

Fraser, N. J., H. Hashimoto and R. J. Cogdell. 2001. Carotenoids and bacterial photosynthesis: The story so far… Photosynthesis Research 70:249-256.

Fuchs, B. M., M. V. Zubkov, K. Sahm, P. H. Burkill, and R. Amann. 2000. Changes in community composition during dilution cultures of marine bacterioplankton as assessed by flow cytometric and molecular biological techniques. Environmental Microbiology 2:191-201.

Fuhrman, J. 1981. Influence of method on the apparent size distribution of Bacterioplankton cells: Epifluoresence microscopy compared to scanning electron microscopy. Marine Ecology Progress Series 5:103-106.

Fuhrman, J. A. 1999. Marine viruses and their biogeochemical and ecological effects. Nature 399:541- 548.

Fuhrman, J. A., and A. A. Davis. 1997. Widespread archaea and novel bacteria from the deep sea as shown by 16S rRNA gene Sequences. Marine Ecology Progress Series 150:275-285.

204 Fuhrman, J. A., and R. T. Noble. 1995. Viruses and Cause Similar Bacterial Mortality in Coastal Seawater. Limnology and Oceanography 40:1236-1242.

Fuhrman,J.A.,K.McCallum,andA.A.Davis.1992. Novel major archaeobacterial group from marine plankton. Nature 356:148-149.

Fuhrman, J. A., and C. C. Ouverney. 1998. Marine microbial diversity studied via 16S rRNA sequences: cloning results from coastal waters and counting of native archaea with fluorescent single cells probes. Aquatic Ecology 32:3-15.

Fuhrman, J. A., T. D. Sleeter, C. A. Carlson, and L. M. Proctor. 1989. Dominance of bacterial biomass in the Sargasso Sea and its ecological implications. Marine Ecology Progress Series 57:207-217.

Fuhrman, J. A., and C. A. Suttle. 1993. Viruses in marine planktonic systems. Oceanography 6:51-63.

Fuller, N. J., N. J. West, D. Marie, M. Yallop, T. Rivlin, A. F. Post and D. J. Scanlan 2005. Dynamics of community structure and phosphate status of picocyanobacterial populations in the Gulf of Aqaba, Red Sea. Limnology and Oceanography 50: 363-375.

Futai, M., H. Omote, Y. Sambongi, and Y. Wada. 2000. Synthase (H+ ATPase): coupling between catalysis, mechanical work, and proton translocation. Biochimica et Biophysica Acta 1458:276-288.

Gasol,J.M.,U.L.Zweifel,F.Peters,J.A.Fuhrman,andÅ.Hagström.1999. Significance of size of nucleic acid content heterogeneity as measured by flow cytometry in natural planktonic bacteria. Applied and Environmental Microbiology 65:4475-4483.

Giometti, C. S., M. A. Gemmel, S. L. Tollaksen, and J. Taylor. 1991. Quantitation of human leukocyte proteins after silver staining: a study with two-dimensional electrophoresis. Electrophoresis 12:536-543.

Giovannoni, S. J., T. B. Britschgi, C. L. Moyer, and K. G. Field. 1990. Genetic diversity in Sargasso Sea bacterioplankton. Nature 345:60-63.

Giovannoni, S. J., and M. Rappé. 2000. Evolution, diversity, and molecular ecology of marine prokaryotes, p. 47-84. In D. L. Kirchman (ed.), Microbial ecology of the oceans. Wiley-Liss Inc., New York.

Giovannoni, S. J., H. J. Tripp, S. Givan, M. Podar, K. L. Vergin, D. Baptista, L. Bibbs, J. Eads, T. H. Richardson, M. Noordewier, M. S. Rappe, J. M. Short, J. C. Carrington and E. J. Mathur 2005. Genome streamlining in a cosmopolitan oceanic bacterium. Science 309: 1242-5.

205 Givskov, M., L. Eberl, and S. Molin. 1994. Responses to nutrient starvation in Pseudomonas putida KT2442 : two-dimensional electrophoretic analysis of starvation- and stress-induced proteins. Journal of Bacteriology 176:4816- 4824.

Givskov, M., L. Eberl, S. Moller, L. K. Poulsen, and S. Molin. 1994. Responses to nutrient starvation in Pseudomonas putida KT 2442 : analysis of general cross- protection, cell shape and macromolecular content. Journal of Bacteriology 176:7-14.

Glöckner, F. O., B. M. Fuchs, and R. Amann. 1999. Bacterioplankton compositions of and oceans: a first comparison based on fluorescence in situ hybridization. Applied and Environmental Microbiology 65:3721-3726.

Glöckner, F. O., E. Zaichikov, N. Belkova, L. Denissova, J. Pernthaler, A. Pernthaler, and R. Amann. 2000. Comparative 16S rRNA analysis of lake bacterioplankton reveals globally distributed phylogenetic clusters including an abundant group of . Applied and Environmental Microbiology 66:5053-65.

Godoy, F., M. Vancanneyt, M. Martínez, A. Steinbuchel, J. Swings and B. H. A Rehm. 2003. Sphingopyxis chilensis sp. nov., a chlorophenol-degrading bacterium which accumulates polyhydroxyalkanoate, and transfer of Sphingomonas alaskensis to Sphingopyxis alaskensis comb. nov. International Journal of Systematic and Evolutionary Microbiology 53: 473-477.

Golovlev, E. L. 1999. An introduction to the biology of the stationary-phase of bacteria: the mechanism of the common response to stresses. Microbiology 68:543-550.

Gooley, A. A., and N. H. Packer. 1997. The importance of protein co- and post- translational modifications in proteome projects, p. 65-89. In M. R. Wilkins, K. L. Williams, R. D. Appel, and D. F. Hochstrasser (ed.), Proteome research: new frontiers in functional genomics. Springer-Verlag, Berlin, Germany.

Goto, S., K. Kogure, K. Abe, Y, Kimata, K. Kitahama, E. Yamashita and H. Terada. 2001. Efficient radical trapping at the surface and inside the phospholipids membrane is responsible for highly potent antiperoxidative activity of the carotenoid astaxanthin. Biochimica Biophysica Acta 1512:251- 258.

Gottschal, J. C. 1990a. Differrent types of continuous cultures for ecological studies. In J. R. N. a. R. Grigorova (ed.), Methods in Microbiology, vol. 22. Acdemic Press, London.

Gottschal, J. C. 1990b. Phenotypic responses to environmental changes. FEMS Microbiology Ecology 74:93-102.

206 Götz, T., U. Windhövel, P. Böger, and G. Sandmann. 1999. Protection of photosynthesis against ultraviolet-B radiation by carotenoids in transformants of the cyanobacterium Synechococcus PCC7942. Plant Physiology 120:599- 604.

Gourmelon, M., J. Cillard, and M. Pommepuy. 1994. Visible light damage to Escherichia coli in seawater: Oxidative stress hypothesis. Journal of Applied Bacteriology 77:105-112.

Gregori,G.,S.Citterio,A.Ghiani,M.Labra,S.Sgorbati,S.Brown,andM. Denis. 2001. Resolution of viable and membrane-compromised bacteria in freshwater and marine waters based on analytical flow cytometry and nucleic acid double staining. Applied and Environmental Microbiology 27:4662-4670.

Guixa-Boixereu, N., D. Vaqué, J. M. Gasol, and C. Pedrós-Alió. 1999. Distribution of viruses and their potential effect on bacterioplankton in an oligotrophic marine system. Aquatic Microbial Ecology 19:205-213.

Gygi,S.P.,G.Corthals,Y.Zhang,Y.Rochon,andR.Aebersold.2000a. Evaluation of two-dimensional gel electrophoresis based proteome analysis technology. Proceedings of the National Academic of Science USA 97:9390- 9395.

Gygi, S. P., D. K. M. Han, A. Gingras, N. Sonenberg, and R. Aebersold. 1999. Protein analysis by mass spectrophotometry and sequence database searching: tools for cancer research in the post-genomic era. Electrophoresis 20:310-319.

Gygi, S. P., B. Rist, and R. Aebersold. 2000b. Measuring gene expression by quantitative proteome analysis. Current Opinion in Microbiology 11:396-401.

Hagström, Å., J. Pinhassi, and U. L. Zweifel. 2000. Biogeographical diversity among marine bacterioplankton. Aquatic Microbial Ecology 21:231-244.

Harder, W., and L. Dijkhuizen. 1983. Physiological responses to nutrient limitation. Annual Review of Microbiology 37:1-23.

Harvey, R. J. 1970. Metabolic regulation in glucose-limited chemostat cultures of Escherichia coli. Journal of Bacteriology 104:698-706.

Hassan, H. M., and I. Fridovich. 1977. Physiological function of superoxide dismutatse in glucose-limited chemostat cultures of E. coli. Journal of Bacteriology 130:805-811.

Hedman, M., H. Deloof, G. Von Heijne and A. Elofsson. 2002. Improved detection of homologous membrane proteins by inclusion of information from topology predictions. Protein Science 11:652-658

207 Heikes, B., M. H. Lee, D. Jacob, R. Talbot, J. Bradshaw, H. Singh, D. Blake, B. Anderson, H. Fuelberg, and A. M. Thompson. 1996. Ozone, hydroperoxides, oxides of nitrogen, and hydrocarbon budgets in the marine boundary layer over the South Atlantic. Journal of Geophysical Research Atmospheres 101:24221-24234.

Hengge-Aronis, R. 1996. Regulation of gene expression during entry into stationary phase, p. 1497-1512. In F.C.Neidhardt,R.CurtissIII,J.L.Ingraham,E.C.C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella : Cellular and Molecular Biology, 2nd ed, vol. 1. American Society for Microbiology, Washington, D.C.

Hengge-Aronis, R. 2002. Signal Transduction and regulatory mechanisms involved in control of the S (RpoS) Subunit of RNA Polymerase. Microbiology and Molecular Biology Reviews 66:373-395.

Herbert, B. R., J. Sanchez, and L. Bini. 1997. Two-dimensional electrophoresis: the state of the art and future directions, p. 13-33. In M. R. Wilkins, K. L. Williams, R. D. Appel, and D. F. Hochstrasser (ed.), Proteome Research: New Frontiers in Functional Genomics. Springer-Verlag, Berlin.

Hirsch, P., M. Bernhard, S. S. Cohen, J. C. Ensign, H. W. Jannasch, A. L. Koch, K. C. Marshall, J. S. Poindexter, S. C. Rittenberg, D. C. Smith, and H. Veldkamp. 1979. Presented at the Strategies of Microbial Life in Extreme Environments, Berlin.

Hoben, H. S. P. 1982. Comparison of the pour, spread and drop plate methods for the enumeration of Rhizobium spp. in inoculants made from presterilized peats. Applied and Environmental Microbiology 44:1246-1247.

Holmquist, L., and S. Kjelleberg. 1993. Changes in viability, respiratory activity and morphology of the marine Vibrio sp strain S14 during starvation of individual nutrients and subsequent recovery. FEMS Microbiology Ecology 12:215-224.

Holms, H. 1996. Flux analysis and control of the central metabolic pathways in Escherichia coli. FEMS Microbiology Reviews 19:85-116.

Hood, M. A., and M. T. MacDonell. 1987. Distribution of ultramicrobacteria in a Gulf Coast Estuary and induction of ultramicrobacteria. Microbial Ecology 14:113-127.

Hoshino, T., R. Fujii and T. Nakahara. 1993. Molecular cloning and sequence analysis of the crtB gene of thermophilus HB27, an extreme producing carotenoid pigments. Applied and Environmental Microbiology. 59:3150-3153.

208 Huang, F., I. Parmryd, F. Nilsson, A. L. Persson, H. B. Pakrasi, B. Andersson and B. Norling. 2002. Proteomics of Synechocystis sp. strain PCC 6803: identification of plasma membrane proteins. Molecular and Cellular Proteomics 1: 956-66.

Huss, V. A. R., H. Festl, and K. H. Schleifer. 1983. Studies on spectrophotometric determination of DNA hybridization from renaturation rates. Systematic and Applied Microbiology 4:184-192.

Ihssen, J. and T. Egli. 2004. Specific growth rate and not cell density controls the general stress response in Escherichia coli. Microbiology 150:1637-1648.

Ishida, Y., and H. Kadota. 1981. Growth patterns and substrate requirements of naturally occurring obligate oligotrophs. Microbial Ecology 7:123-130.

Jannasch, H. W. a. G. E. Jones. 1959. Bacterial populations in seawater as determined by different methods of enumeration. Limnology and Oceanography 4:128-139.

Janssen, P. H., A. Schuhmann, E. Morschel, and F. A. Rainey. 1997. Novel anaerobic ultramicrobacteria belonging to the Verrucomicrobiales lineage of bacterial descent isolated by dilution culture from anoxic rice paddy soil. Applied and Environmental Microbiology 63:1382-1388.

Jenkins, C. L., A. G. Andrewes, T. J. McQuade, and M. P. Starr. 1979. The pigment of Pseudomonas paucimobilis is a carotenoid (Nostoxanthin), rather than a brominated aryl-polyene (Xanthomonadin). Current Microbiology 3:1-4.

Jenkins, D. E., S. A. Chaisson, and A. Matin. 1990. Starvation-induced cross protection against osmotic challenge in Escherichia coli. Journal of Bacteriology. 172:2779-2781.

Jenkins, D. E., J. E. Schultz, and A. Matin. 1988. Starvation-induced cross protection against heat or H2O2 challenge in Escherichia coli. Journal of Bacteriology. 170:3910-3915.

Jouper-Jaan, A., A. E. Goodman, and S. Kjelleberg. 1992. Bacteria starved for prolonged periods develop increased protection against lethal temperatures. FEMS Microbiology Reviews 101:229-236.

Joux, F., W. H. Jeffrey, P. Lebaron, and D. L. Mitchell. 1999. Marine bacterial isolates display diverse responses to UV-B radiation. Applied and Environmental Microbiology 65:3820-7.

Joux, F., and P. Lebaron. 2000. Use of fluorescent probes to assess physiological functions of bacteria at single-cell level. Microbes and Infection 2:1523-35.

Jungblut, P. R., D. Bumann, G. Haas, U. Zimny-Arndt, P. Holland, S. Lamer, F. Siejak, A. Aebischer, and T. F. Meyer. 2000. Comparative proteome analysis of Helicobacter pylori. Molecular Microbiology 36:710-725.

209 Jürgens, K., J. M. Gasol, and D. Vaqué. 2000. Bacteria-flagellate coupling in microcosm experiments in the Central Atlantic Ocean. Journal of Experimental Marine Biology and Ecology 245:127- 147.

Jürgens, K., J. Pernthaler, S. Schalla, and R. Amann. 1999. Morphological and compositional changes in a planktonic bacterial community in response to enhanced protozoan grazing. Applied and Environmental Microbiology 65:1241-50.

Karner, M., and J. A. Fuhrman. 1997. Determination of active marine bacterioplankton: a comparison of universal 16S rRNA probes, autoradiography, and nucleoid staining. Applied and Environmental Microbiology 63:1208-1213.

Keil, R. G., and D. L. Kirchman. 1986. Dissolved Combined Amino Acids - Chemical Form and Utilization By Marine Bacteria. Limnology and Oceanography 38:1256-1270.

Keil, R. G., and D. L. Kirchman. 1991. Contribution of dissolve free amino acids and ammonium to the nitrogen requirements of heterotrophic bacterioplankton. Marine Ecology Progress Series 73:1-10.

Keil, R. G., and D. L. Kirchman. 1999. Utilization of dissolved protein and amino acids in the northern Sargasso Sea. Aquatic Microbial Ecology 18:293-300.

Kell, D.B., A. S. Kaprelyants, D. H. Weichart, C. R. Harwood, M. R. and Barer. 1998 Viability and activity in readily culturable bacteria: a review and discussion of the practical issues. Antonie Van Leeuwenhoek 73:169-187.

Kemp, P. F., S. Lee, and J. Laroche. 1993. Estimating the growth rate of slowly growing marine bacteria from RNA content. Applied and Environmental Microbiology 59:2594-2601.

Kirchman, D. L. 1990. Limitation of bacterial growth by dissolved organic matter in the subarctic Pacific. Marine Ecology Progress Series 62:47-54.

Kirchman, D. L. 1994. The Uptake of Inorganic Nutrients By Heterotrophic Bacteria. Microbial Ecology 28:255-271.

Kirchman, D. L., and R. E. Hodson. 1986. Metabolic regulation of amino acid uptake in marine waters. Limnology and Oceanography 31:339-350.

Kirchman, D. L., R. G. Keil, M. Simon, and N. A. Welschmeyer. 1993. Biomass and Production of Heterotrophic Bacterioplankton in the Oceanic Subarctic Pacific. Deep Sea Research Part I Oceanographic Research Papers 40:967-988.

Kirchman, D. L., B. Meon, M. T. Cottrell, D. A. Hutchins, D. Weeks, and K. W. Bruland. 1981. Carbon versus iron limitation of bacterial growth in the California upwelling regime. Limnology and Oceanography 45:1681-1688.

210 Kirchman, D. L., and J. H. Rich. 1997. Regulation of Bacterial Growth Rates By Dissolved Organic Carbon and Temperature in the Equatorial Pacific Ocean. Microbial Ecology 33:11-20.

Kirchman,D.L.,J.H.Rich,andR.T.Barber.1995. Biomass and biomass production of heterotrophic bacteria along 140˚W in the Equatorial Pacific: effect of temperature on the . Deep-Sea Research II 42:603-619.

Kirchman, D. L., and P. A. Wheeler. 1998. Uptake of Ammonium and Nitrate By Heterotrophic Bacteria and Phytoplankton in the Sub-Arctic Pacific. Deep Sea Research Part I Oceanographic Research Papers 45:347-365.

Kjelleberg, S., N. Albertson, K. Flärdh, L. Holmquist, Å. Jouper-Jaan, R. Marouga, J. Östling, B. Svenblad, and D. Weichart. 1993. How do non- differentiating bacteria adapt to starvation. Antonie van Leeuwenhoek 63:333- 341.

Kjelleberg, S., and M. Hermansson. 1984. Starvation-induced effects on bacterial surface characteristics. Applied and Environmental Microbiology 48:497-506.

Kjelleberg, S., M. Hermansson, P. Mården, and G. W. Jones. 1987. The transient phase between growth and non-growth of heterotrophic bacteria, with emphasis on the marine environment. Annual Review in Microbiology 41:25- 49.

Koch,A.L.1979. Presented at the Strategies of Microbial Life in Extreme Environments, Berlin.

Konstantinidis, K. T. and J. M. Tiedje. 2005. Genomic insights that advance the species definition for prokaryotes. Proceedings of the National Academy of Science USA 102:2567-72.

Lebaron, P., P. Servais, H. Agogue, C. Courties, and F. Joux. 2001. Does the high nucleic acid content of individual bacterial cells allow us to discriminate between active cells and inactive cells in aquatic systems? Applied and Environmental Microbiology 67:1775-82.

Lee, S., and J. A. Fuhrman. 1987. Relationships between biovolume and biomass of naturally derived marine bacterioplankton. Applied and Environmental Microbiology 53: 1298-1303.

Lee, N., P. H. Nielsen, K. H. Andreasen, S. Juretschko, J. L. Nielsen, K. H. Schleifer, and M. Wagner. 1999. Combination of in situ hybridization and microautoradiography- a new tool for structure-function analyses in microbial ecology. Applied and Environmental Microbiology 65:1289-1297.

Lester, P. and S. J. Hubbard. 2002. Comparative bioinformatic analysis of complete proteomes and protein parameters for cross-species identification in proteomics. Proteomics 2:1392-1405.

211 Lewin, B. 1990. Genes IV. Oxford University Press, New York.

Li, W. 1994. Primary Production of Prochlorophytes, Cyanobacteria, and Eukaryotic Ultraphytoplankton: Measurements From Flow Cytometric Sorting. Limnology and Oceanography 39: 169-175.

Li, W. K. W., P. M. Dickie, I. B. D., and A. M. Wood. 1992. Biomass of bacteria, cyanobacteria, prochlorophytes and photosynthetic eukaryotes in the Sargasso Sea. Deep-Sea Research I 39:501-519.

Liaaen-Jensen, S. and A. Jensen. 1971. Quantitative determination of carotenoids in photosynthetic tissue. In: San Pietro, A. (Ed). Methods in Enzymology. Vol 2. Academic press, Orlando, pp586-602.

Liska, A. J. and A. Shevchenko. 2003. Expanding the organismal scope of proteomics: cross-species protein identification by mass spectrometry and its implications. Proteomics 3:19-28.

Liu, X., and T. Ferenci. 1998. Regulation of porin-mediated outer membrane permeability by nutrient limitation in Escherichia coli. Journal of Bacteriology 180:3917-3922.

Maalöe, O. and N. O. Kjeldgaard. 1966. Control of Macromolecular Synthesis. New York: W.A. Benjamin.

MacDonell, M. T., and L. Hood. 1982. Isolation and characterization of ultramicrobacteria from a Gulf Coast estuary. Applied and Environmental Microbiology 45:566-571.

MacGregor, B. J. 1999. Molecular approaches to the study of aquatic microbial communities. Current Opinion in Biotechnology 12:220-224.

Maniatis, T., J. Sambrook, and E. F. Fritsch. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbour Laboratory Press, New York.

Mården, P., T. Nyström, and S. Kjelleberg. 1987. Uptake of leucine by a marine gram-negative bacterium during exposure to starvation conditions. FEMS Microbiology Ecology 45:233-241.

Mården, P., A. Tunlid, K. Malmcrona-Friberg, G. Odham, and S. Kjelleberg. 1985. Physiological and morphological changes during short term starvation of marine bacterial isolates. Archives in Microbiology 142:326-332.

Martin, P., and R. A. MacLeod. 1984. Observations on the distinctions between oligotrophic and eutrophic marine bacteria. Applied and Environmental Microbiology 47:1017-1022.

Matear, R. J., and A. C. Hirst. 1999. Climate change feedback on the future oceanic CO2 uptake. Tellus 51:722-733.

212 Matin, A. 1981. Regulation of enzyme synthesis as studied in continuous culture., p. p 69-97. In P. H. Calcott (Ed.), Continuous Culture of Cells, vol. 2. CRC Press, Boca Raton, Fla.

Matin, A. 1990. Molecular analysis of the starvation stress in Escherichia coli. FEMS Microbiology Ecology 74:185-196.

Matin, A., A. Grootjans, and H. Hogenhuis. 1976. Influence of dilution rate on enzymes of intermediary metabolism. Journal of General Microbiology 94:323-332.

Matin, A., and M. K. Matin. 1982. Cellular levels, excretion, and synthesis rates of cyclic AMP in Escherichia coli grown in continuous culture. Journal of Bacteriology. 149:801-807.

Matin, A. A., and H. Veldkamp. 1978. Physiological basis of the selective advantage of a Spirillum sp. in a carbon-limited environment. Journal of General Microbiology 105:187-197.

Merril, C. R., D. Goldman, and M. L. Van Keuren. 1984. Gel protein stains: silver stain. Methods in Enzymology 104:441-447.

Miller, J. H. 1992. A short course in bacterial genetics: a laboratory manual and handbook for Escherichia coli and related bacteria. Cold Spring Harbor Laboratory. Cold Spring Harbour.

Mitchell, D.L, J. Meador, L. Paniker, D. Gasparutto, W. H. Jeffrey, and J. Cadet. 2002. Development and application of a novel immunoassay for measuring oxidative DNA damage in the environment. Photochemistry and Photobiology. 75: 257-265.

Moaledj, K. 1978. Qualitative assessment of oligocarbophilicaquatic microflora in the Plußee. Archives of hydrobiology 82:98-113.

Molloy,M.P.,N.D.Phadke,J.R.Maddock,andP.C.Andrews.2001. Two- dimensional electrophoresis and peptide mass fingerprinting of bacterial outer membrane proteins. Electrophoresis 22:1686-1696.

Monger, B. C., and M. R. Landry. 1993. Flow cytometric analysis of marine bacteria with Hoechst 33342. Marine Ecology Progress Series 59:908-911.

Moore, L. R., and S. W. Chisholm. 1999. Photophysiology of the marine cyanobacterium Prochlorococcus: Ecotypic differences among cultured isolates. Limnology and Oceanography 44:628-638.

Moore, L. R., R. Goericke, and S. W. Chisholm. 1995. Comparative physiology of Synechococcus and Prochlorococcus - influence of light and temperature on growth, pigments, fluorescence and absorptive properties. Marine Ecology Progress Series 116:259-275.

213 Moore, L. R., M. Ostrowski, D. J. Scanlan, K. Feren and T. Sweetsir. 2005. Ecotypic variation in phosphorus acquisition mechanisms within marine picocyanobacteria. Aquatic Microbial Ecology 39:257-269.

Moore, L. R., G. Rocap, and S. W. Chisholm. 1998. Physiology and molecular phylogeny of coexisting Prochlorococcus ecotypes. Nature 393:464-467.

Moran, M. A., and R. G. Zepp. 1997. Role of photoreactions in the formation of biologically labile compounds from dissolved organic matter. Limnology and Oceanography 42:1307-1316.

Morel, A., Y.-W. Ahn, F. Partensky, D. Vaulot, and H. Claustre. 1993. Prochlorococcus and Synechococcus: a comparative study of their size, pigmentation and related optical properties. Journal of Marine Research 51:617-649.

Moriarty,D.J.W.,andR.T.Bell.1993. Bacterial Growth and Starvation in Aquatic Environments, p. 25-53. In K. S (Ed.), Starvation in Bacteria. Plenum Press, New York.

Morita, R. Y. 1982. Starvation survival of heterotrophs in the marine environment. Advances in Microbial Ecology 61:171-198.

Morita R. Y.1988. Bioavailability of energy and its relationship to growth and starvation survival in nature. Canadian Journal of Microbiology 34: 436-441.

Morita, R. Y. 1985. Starvation and miniaturisation of heterotrophs, with emphasis of maintenance of the starved viable state, p. 111-130. In M. Fletcher and G. Floodgate (Ed.), Bacteria in their : the effect of nutrient conditions. Academic Press, New York, N.Y.

Morita, R. Y. 1997. Physiology of starvation-survival, p. 274-350. In R. Y. Morita (ed.), Bacteria in oligotrophic environments, starvation-survival lifestyle. Chapman and Hall, New York.

Morris,R.M.,M.S.Rappe,S.A.Connon,K.L.Vergin,W.A.Siebold,C.A. Carlson and S. J. Giovannoni 2002. SAR11 clade dominates ocean surface bacterioplankton communities. Nature 420: 806-10.

Moyer, C. L., and R. Y. Morita. 1989a. Effect of growth rate and starvation- survival on the viability and stability of a psychrophilic marine bacterium. Applied and Environmental Microbiology 55:1122-1127.

Moyer, C. L., and R. Y. Morita. 1989b. Effect of growth rate and starvation- survival on cellular DNA, RNA and protein of a psychrophilic marine bacterium. Applied and Environmental Microbiology 55:2710-2716.

Münster, U. 1993. Concentration and fluxes of organic carbon substrates in the aquatic environment. Antonie van Leeuwenhoek 63:243-274.

214 Murray, M. G. andW.F.Thompson.1980. Rapid isolation of high-molecular- weight plant DNA. Nucleic Acids Research 8:4321-4325.

Nagata, T., and D. L. Kirchman. 1997. Roles of submicron particles and colloids in microbial food webs and biogeochemical cycles within marine environments. Advances in Microbial Ecology 15:81-103.

Neef A., R. Witzenberger and P. Kampfer. 1999. Detection of sphingomonads and in situ identification in activated sludge using 16S rRNA-targeted oligonucleotide probes. Journal of Industrial Microbiology and Biotechnology. 23:261-267

Neidhardt, F. C., J. L. Ingraham, and M. Schaechter. 1990. Physiology of the Bacterial Cell: a Molecular Approach. Sinauer Ass. Inc., Massachucetts.

Neilan, B. A., D. Jacobs, T. Del Dot, L. L. Blackall, P. R. Hawkins, P. T. Cox, and A. E. Goodman. 1997. rRNA sequences and evolutionary relationships among toxic and non-toxic cyanobacteria of the genus Microcystis. International Journal of Systematic Bacteriology 47:693-697.

Nelson, D. R., Y. Sadlowski, M. Eguchi, and S. Kjelleberg. 1997. The starvation- stress response of Vibrio (Listonella) anguilarum. Microbiology 143:2305- 2312.

Nichols, P. D., J. B. Guckert, and D. C. White. 1986. Determination of monounsaturated fatty acid double-bond position and geometry for microbial monocultures and complex consortia by capillary GC-MS of their dimethyl disulphide adducts. Journal of Microbiology Methods 5:49-55.

Nichols, D. S., T. A. McMeekin, and P. D. Nichols. 1994. Manipulation of polyunsaturated, branched-chain and trans-fatty acid production in Shewanella putrefaciens strain ACAM 342. Microbiology 140:577-584.

Nierman, W. C., T. V. Feldblyum, M. T. Laub, I. T. Paulsen, K. E. Nelson, J. Eisen,J.F.Heidelberg,M.R.K.Alley,N.Ohta,J.R.Maddock,I. Potocka, W. C. Nelson, A. Newton, C. Stephens, N. D. Phadke, B. Ely, R. T. DeBoy, R. J. Dodson, A. S. Durkin, M. L. Gwinn, D. H. Haft, J. F. Kolonay, J. Smit, M. B. Craven, and H. Khouri. 2001. Complete genome sequence of Caulobacter crescentus. Proceedings of the National Academic of Science USA 98:4136-4141.

Noji, H., and M. Yoshida. 2000. The rotary machine in the cell, ATP synthase. Journal of Biological Chemistry 276:1665-1668.

Notley, L., and T. Ferenci. 1995. Differential expression of mal genes under cAMP and endogeneous inducer control in nutrient stressed Escherichia coli. Molecular Microbiology 160:121-129.

215 Notley, L., and T. Ferenci. 1996. Induction of RpoS-dependent functions in glucose-limited continuous culture: what level of nutrient limitation induces the stationary phase of Escherichia coli? Journal of Bacteriology 178:1465- 1468.

Notley-McRobb, L., A. Death, and T. Ferenci. 1997. The relationship between external glucose concentration and cAMP levels inside Escherichia coli: implications for models of phosphotransferase-mediated regulation of adenylate cyclase. Microbiology 143:1909-1918.

Notley-McRobb, L. and T. Ferenci. 1999. Adaptive mgl-regulatory mutations and genetic diversity evolving in glucose-limited Escherichia coli populations. Environmental Microbiology 1: 33-43.

Notley-McRobb, L. and T. Ferenci 1999. The generation of multiple co-existing mal-regulatory mutations through polygenic evolution in glucose-limited populations of Escherichia coli. Environmental Microbiology 1: 45-52.

Nouwens, A.S., S. J. Cordwell, M. R. Larsen, M. P. Molloy, M. Gillings , M. D. Willcox and B. J. Walsh. 2000. Complementing genomics with proteomics: the membrane subproteome of Pseudomonas aeruginosa PAO1. Electrophoresis 21:3797-809.

Nouwens, A. S., F. G. Hopwood, M. Traini, K. L. Williams, and B. J. Walsh. 1999. Proteome approach to the identification of cellular Escherichia coli proteins, p. 331-349. In R. L. Charlebois (Ed.), Organization of the prokaryotic genome. American Society for Microbiology, Washington D.C.

Novitsky,J.A.,andR.Y.Morita.1976. Morphological characterization of small cells resulting from nutrient starvation of a psychrophilic marine Vibrio. Applied and Environmental Microbiology 32:617-622.

Nyström, T. 1994. The glucose-starvation stimulon of Escherichia coli: induced and repressed synthesis of enzymes of central metabolic pathways and role of acetyl phosphate in gene expression and starvation survival. Molecular Microbiology 12:833-843.

Nyström, T. 1999. Starvation, cessation of growth and bacterial aging. Current Opinions in Microbiology 2:214-219.

Nyström, T. 2003 Nonculturable bacteria: programmed survival forms or cells at death's door? Bioessays 25:204-211.

Nyström, T., N. H. Albertson, K. Flardh, and S. Kjelleberg. 1990b. Physiological and Molecular Adaptation to Starvation and Recovery from Starvation by the Marine Vibrio sp. S14. FEMS Microbial Ecology 74:129-140.

Nyström, T., K. Flärdh, and S. Kjelleberg. 1990a. Responses to multiple-nutrient starvation in marine Vibrio sp. strain CCUG 15956. Journal of Bacteriology 172:7085-7097.

216 Nyström, T., and S. Kjelleberg. 1989. Role of protein synthesis in the cell division and starvation induced resistance to autolysis of a marine Vibrio during the initial phase of starvation. Journal of General Microbiology 135:1599-1606.

Nyström,T.,P.Mården,andS.Kjelleberg.1986. Relative changes in incorporation rates of leucine and methionine during starvation survival of two bacteria isolated from marine waters. FEMS Microbiology Ecology 38:285- 292.

Nyström, T., R. M. Olsson, and S. Kjelleberg. 1992. Survival, stress resistance and alterations in protein expression in the marine Vibrio sp. strain S14 during starvation for different individual nutrients. Applied and Environmental Microbiology 58:55-65.

Oda,Y.,S.J.Slagman,W.G.Meijer,L.J.Forney,andJ.C.Gottschal.2000. Influence of growth rate and starvation on fluorescent in situ hybridization of Rhodopseudomonas palustris. FEMS Microbiology Ecology 32:205-213.

O'Farell, P. H. 1975. High resolution two-dimensional electrophoresis of protein. Journal of Biological Chemistry 250:4007-4021.

Oosthuizen, M. C., B. Steyn, J. Theron, P. Cosette, D. Lindsay, D., A. von Hoply and V. S. Brözel. (2002) Proteomic analysis reveals differential protein expression by Bacillus cereus during biofilm formation. Applied and Environmental Microbiology 68: 2770-2780.

Östling, J. 1996. On the regulation of the carbon starvation response in a marine Vibrio sp. PhD thesis. Goteborg University, Goteborg.

Ourisson, G., and Y. Natatani. 1990. Bacterial carotenoids as membrane reinforcers. A general role for polyterpenoids: Membrane stabilisation. In I. Kinsky (ed.), Carotenoids: Chemistry and Biology. Plenum Press, New York.

Ouverney, C. C., and J. A. Fuhrman. 1999. Combined Microautoradiography-16S rRNA probe technique for determination of radioisotope uptake by specific microbial cell types in situ. Applied and Environmental Microbiology 65:1746-1752.

Ouverney, C. C. and J. A. Fuhrman 2000. Marine planktonic archaea take up amino acids. Applied and Environmental Microbiology 66: 4829-33.

Papayannopoulos, I. A. 1995. The interpretation of collision-induced dissociation tandem mass spectra of peptide. Mass Spectrometry Reviews 14:49-73.

Pappin, D. J. 1997. Peptide mass fingerprinting using MALDI-TOF mass spectrometry. Methods in Molecular Biology 64:165-173.

217 Pappin, D. J. C., P. Hojrup, and A. J. Bleasby. 1993. Rapid identification of proteins by peptide-mass fingerprinting. Current Biology 3:327-332.

Parpais, J., Marie, D., Partensky, F., Morin, P., and Vaulot, D. 1996 Effect of Phosphorus Starvation On the Cell-Cycle of the Photosynthetic Prokaryote Prochlorococcus Spp. Marine Ecology-Progress Series 132: 265-274.

Partensky,F.,J.Blanchot,andD.Vaulot.1999b. Differential distribution and ecology of Prochlorococcus and Synechococcus in oceanic waters: a review. Bulletin of the Institute of Oceanography. Monaco. n. esp 19:431-449.

Partensky,F.,W.R.Hess,andD.Vaulot.1999a. Prochlorococcus,amarine photosynthetic prokaryote of global significance. Microbiology and Molecular Biology Reviews 63:106-127.

Paulsen, I. T., M. K. Sliwinski, and M. H. Saier. 1998. Microbial genome analysis: global comparisons of transport capabilities based on phylogenies, bioenergetics and substrate specificities. Journal of Molecular Biology 277:573-592.

Pedrós-Alió,C.,andT.D.Brock.1983. The importance of attachment particles for planktonic bacteria. Archiv Fuer Hydrobiologie 98:354- 379.

Pernthaler, A., J. Pernthaler, H. Eilers, and R. Amann. 2001. Growth patterns of two marine isolates: adaptations to substrate patchiness? Applied and Environmental Microbiology 67:4077-83.

Pernthaler, A., C. M. Preston, J. Pernthaler, E. F. DeLong, and R. Amann. 2002. Comparison of fluorescently labelled oligonucleotide and polynucleotide probes for the detection of pelagic marine bacteria and archaea. Applied and Environmental Microbiology 68:661-7.

Pinhassi, J., and A. Hagstrom. 2000. Seasonal succession in marine bacterioplankton. Aquatic Microbial Ecology 21:245-256.

Pinhassi, J., F. Azam, J. Hemphala, R. A. Long, J. Martinez, U. L. Zweifel, and A. Hagstrom. 1999. Coupling between bacterioplankton species composition, , and organic matter degradation. Aquatic Microbial Ecology 17:13-26.

Pinhassi, J., U. L. Zweifel, and A. Hagstrom. 1997. Dominant Marine Bacterioplankton Species Found Among Colony-Forming Bacteria. Applied and Environmental Microbiology 63:3359-3366.

Pirt,S.J.1965. The maintenance energy of bacteria in growing cultures. Proceedings of the Royal Society of London Series B 163:223-231.

Pirt,S.J.1982. Maintenance energy: a general model for energy-limited and energy-sufficient growth. Archives of Microbiology 133:300-2.

218 Poindexter, J. S. 1979. Presented at the Strategies of Microbial Life in Extreme Environments, Berlin.

Poindexter, J. S. 1981. Oligotrophy: feast and famine existence. Advances in Microbial Ecology 5:63-89.

Pollock,T.J.,andR.W.Armentrout.1999. Planktonic/sessile dimorphism of polysaccharide-encasulated sphingomonads. Journal of Industrial Microbiology and Biotechnology 23:436-441.

Poolman, B., E. J. Smid, H. Veldkamp , and W. N. Konings. 1987. Bioenergetic consequences of lactose starvation in contiuously cultured Streptococcus cremoris. Journal of Bacteriology. 169:1460-1468.

Porter, K. G., and Y. S. Feig. 1980. The use of DAPI for identifying and counting the aquatic microflora. Limnology and Oceanography 25:943-948.

Preyer, J. M. and J. D. Oliver. 1993. Starvation-induced thermal tolerance as a survival mechanism in a marine bacterium. Applied and Environmental Microbiology 59:2653-2656

Rabilloud, T., L. Vuillard, C. Gilly, and J. J. Lawrence. 1994. Silver-staining of proteins in polyacrylamide gels: a general overview. Cellular and Molecular Biology 40:57-75.

Rappé, M. S., S. A. Connon, K. L. Vergin, and S. J. Giovannoni. 2002. Cultivation of the ubiquitous SAR11 marine bacterioplankton clade. Nature 418:630-3.

Rappé, M. S., K. Vergin, and S. J. Giovannoni. 2000. Phylogenetic comparisons of a coastal bacterioplankton community with its counterparts in open ocean and freshwater systems. FEMS Microbiology Ecology 33:219-232.

Righetti, P. G. 1990. Immobilised pH gradients: theory and methodology. In R. H. Burdon and P. H. van Knippenberg (Ed.), Laboratory techniques in biochemistry and molecular biology. Elsevier, Amsterdam.

Robertson,B.R.,D.K.Button,andA.L.Koch.1998. Determination of the biomasses of small bacteria at low concentrations in a mixture of species with forward light scatter measurements by flow cytometry. Applied and Environmental Microbiology 64:3900-3909.

Rocap,G.,F.W.Larimer,J.Lamerdin,S.Malfatti,P.Chain,N.A.Ahlgren,A. Arellano, M. Coleman, L. Hauser, W. R. Hess, Z. I. Johnson, M. Land, D. Lindell, A. F. Post, W. Regala, M. Shah, S. L. Shaw, C. Steglich, M. B. Sullivan, C. S. Ting, A. Tolonen, E. A. Webb, E. R. Zinser and S. W. Chisholm. 2003. Genome divergence in two Prochlorococcus ecotypes reflects oceanic . Nature 424:1042-7.

219 Rørth, M., and P. K. Jensen. 1967. Determination of catalase activity by means of the Clark oxygen electrode. Biochimica et Biophysica Acta 139:173-176.

Rowe, N. J., J. Tunstall, L. Galbraith, and S. G. Wilkinson. 2000. Lipid composition and of Pseudomonas echinoides: transfer to the genus Sphingomonas. Microbiology 146:3007-3012.

Saitou, N., and M. Nei. 1987. The neighbour-joining method: a new method for reconstructing phylogenetic trees. Molecular Biology and Evolution 4:406- 425.

Sambrook,J.E.,E.F.Fritsch,andT.Maniatis.1989. Molecular Cloning: a laboratory manual, 2nd. ed. Cold Spring Harbour Lab. Press, New York.

Sandmann, G. D. S. Kuhn and P. Böger. 1998. Evaluation of structurally different carotenoids in Escherichia coli transformants as protectants against UV-B radiation. Applied and Environmental Microbiology 64:1972-1974.

Sanudo-Wilhelmy, S.A., Kustka, A.B., Gobler, C.J., Hutchins, D.A., Yang, M., Lwiza, K., Burns, J., Capone, D.G., Raven, J.A., and Carpenter, E.J. 2001 Phosphorus limitation of nitrogen fixation by Trichodesmium in the central Atlantic Ocean. Nature 411: 66-69.

Scanlan, D. J., N. J. Silman, K. M. Donald, W. H. Wilson, N. G. Carr, I. Joint and N. H. Mann 1997. An immunological approach to detect phosphate stress in populations and single cells of photosynthetic picoplankton. Appl Environ Microbiol 63:2411-20.

Schultz, J. E., G. I. Latter, and A. Matin. 1998. Differential regulation by cyclic AMP of starvation protein synthesis in Escherichia coli. Journal of Bacteriology 170:3903-3909.

Schut, F. 1993. Ecophysiology of a marine ultramicrobacterium. Ph.D. thesis. University of Groningen, Groningen, The Netherlands.

Schut, F., J. C. Gottschal, and R. A. Prins. 1997a. Isolation and characterisation of the marine ultramicrobacterium Sphingomonas sp strain RB2256. FEMS Microbiology Reviews 20:363-369.

Schut, F., M. Jansen, T. M. Pedro-Gomes, J. C. Gottschal, W. Harder, and R. A. Prins. 1995. Substrate uptake and utilization by a marine ultramicrobacterium. Microbiology 141:351-361.

Schut,F.,R.A.Prins,andJ.C.Gottschal.1997b. Oligotrophy and pelagic marine bacteria : facts and fiction. Aquatic Microbial Ecology 12:177-202.

Schweder, T., A. Kolyschkow, U. Völker and M. Hecker. 1999. Analysis of the expression and function of the B-dependent general stress regulon of Bacillus subtilus during slow growth. Archives of Microbiology 171:439-434.

220 Schut, F., E. J. de Vries, J. C. Gottschal, B. R. Robertson, W. Harder, R. A. Prins, and D. K. Button. 1993. Isolation of typical marine bacteria by dilution culture : growth, maintenance, and characteristics of isolates under laboratory conditions. Applied and Environmental Microbiology 59:2150-2160.

Servais, P., C. Courties, P. Lebaron, and M. Trousseillier. 1999. Coupling bacterial activity with cell sorting by flow cytometry. Microbial Ecology 38:180-189.

Servais, P., H. Agogue, C. Courties, F. Joux, and P. Lebaron. 2001. Are the actively respiring cells (CTC+) those responsible for bacterial production in aquatic environments? FEMS Microbiology Ecology 35:171-179.

Shanmuganathan, A., S. V. Avery, S. A. Willets and J. E. Houghton. 2004. Copper-induced oxidative stress in Saccharomyces cerevisiae targets enzymes of the glycolytic pathway. FEBS Letters 556:253-259

Sherr, E. B., and B. F. Sherr. 1994. Bacterivory and herbivory: key roles of phagotrophic protists in pelagic food webs. Microbial Ecology 28:223-235.

Shevchenko, A., S. Sunyaev, A. Loboda, A. Shevehenko, P. Bork, W. Ens, and K. G. Standing. 2001. Charting the proteomes of organisms with unsequenced genomes by MALDI-quadrupole time of flight mass spectrometry and BLAST homology searching. Analytical Chemistry 73:1917-1926.

Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric sequencing of proteins from silver-stained polyacrylamide gels. Analytical Chemistry 68:850-858.

Sickmann, A., K. Marcus, H. Schafer, E. Butt-Dorje, S. Lehr, A. Herkner, S. Suer, I. Bahr, and H. E. Meyer. 2001. Identification of post-translationally modified proteins in proteome studies. Electrophoresis 22:1669-1676.

Sieburth J. M. N., V. Smetacek and J. Lenz. 1978. Pelagic ecosystem structure: heterotrophic compartments of the plankton and their relationship to plankton size fractions. Limnology and Oceanography 23: 1256-1263.

Sieracki, M. E., E. M. Haugen, and T. L. Cucci. 1995. Overestimation of heterotrophic bacteria in the Sargasso Sea: direct evidence by flow and imaging cytometry. Deep-Sea Research I 42:1399-1409.

Simek, K., and H. Chrzanowski. 1992. Direct and indirect evidence of size- selective grazing on pelagic bacteria by freshwater nanoflagellates. Applied and Environmental Microbiology 58:3715-3720.

Simon, M., and F. Azam. 1989. Protein content and protein synthesis rates of planktonic marine bacteria. Marine Ecology Progress Series 51:201-213.

221 Sly, L. I., L. L. Blackall, P. C. Kraat, T. Tian-Shen, and V. Sangkhobol. (1986). The use of second derivative plots for the determination of mol% guanine plus cytosine of DNA by the thermal denaturation method. Journal of Microbiology Methods. 5:139-156.

Smith, P. K., R. I. Krohn, G. T. Hermanson, A. K. Mallia, F. H. Gartner, M. D. Provenzano, E. K. Fujimoto, N. M. Goeke, B. J. Olson, and D. C. Klenk. 1985. Measurement of protein using bicinchoninic acid. Analytical Biochemistry 150:76-85.

Srinivasan, S., J. Östling, T. Charlton, R. de Nys, K. Takayama, and S. Kjelleberg. 1998. Extracellular signal molecule(s) involved in the carbon starvation response of marine Vibrio sp. strain S14. Journal of Bacteriology 180:201-209.

Stackebrandt, E., and B. M. Goebel. 1994. Taxonomic note: a place for DNA- DNA reassociation and 16S rRNA sequence analysis in the present species definition in bacteriology. International Journal of Systematic Bacteriology. 44:846-849.

Storz, G., and Imlay, J.A. 1999 Oxidative stress. Current Opinion in Microbiology 2: 188-194.

Strehl, B., J. Holtzendorff, F. Partensky, and W. R. Hess. 1999. A small and compact genome in the marine cyanobacterium Prochlorococcus marinus CCMP 1375: lack of an intron in the gene for tRNA(Leu)(UAA) and a single copy of the rRNA operon. FEMS Microbiology Letters 181:261-266.

Subczynski, W. K., E. Markowska, and J. Sielewiesiuk. 1991. Effect of polar carotenoids on the oxygen diffusion-concentration product in lipid bilayers. An EPR spin labell study. Biochimica et Biophysica Acta 1068:68-72.

Tabata, K., K. Kasuya, H. Abe, K. Masuda, and Y. Doi. 1999. Poly(aspartic acid) degradation by a Sphingomonas sp. isolated from freshwater. Applied and Environmental Microbiology 65:4268-4270.

Takeuchi, M., K. Hamana, and A. Hiraishi. 2001. Proposal of the genus Sphingomonas sensu stricto and three new genera, Sphingobium, Novosphingobium and Sphingopyxis, on the basis of phylogenetic and chemotaxonomic analyses. International Journal of Systematic and Evolutionary Microbiology 51:1405-1417.

Teich, A., S. Meyer, H. Y. Lin, L. Andersson, S. O. Enfors, and P. Neubauer. 1999. Growth rate related concentration changes of the starvation response regulators sigma(S) and ppGpp in glucose-limited fed-batch and continuous cultures of Escherichia coli. Biotechnology Progress 15:123-129.

222 Tempest, D. W., D. Herbert, and P. J. Phipps. 1967. Studies on the growth of Aerobacter aerogenes at low dilution rates in a chemostat. pp.240-253. In: Powell, E.O., C.G.T. Evans, R.E. Strange, and D. W. Tempest (Eds.) Microbial Physiology and Continuous Culture.

Tempest, D. W. and O. M. Neijssel. 1978. Eco-physiological aspects of microbial growth in aerobic nutrient-limited environments. Advances in Microbial Ecology 2:105-153.

Thingstad, T. F., and R. Lignell. 1997. Theoretical models for the control of bacterial growth rate, abundance, diversity and carbon demand. Aquatic Microbial Ecology 13:19-27.

Thingstad, T. F., U. L. Zweifel, and F. Rassoulzadegan. 1998. P limitation of heterotrophic bacteria and phytoplankton in the Northwest Mediterranean. Limnology and Oceanography 43:89-94.

Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin and D. G. Higgins. 1997. The ClustalX windows interface: flexible strategies for multipler sequence alignment aided by quality analysis tools. Nucleic Acids Research 24:4876-4882.

Tonella, L., B. J. Walsh, J. Sanchez, K. Ou, M. R. Wilkins, M. Tyler, S. Frutiger, A. A. Gooley, I. Pescaru, R. F. Appel, J. X. Yan, A. Bairoch, C. Hoogland, F. S. Morch, G. J. Hughes, K. L. Williams, and D. F. Hochstrasser. 1998. '98 Escherichia coli SWISS-2DPAGE database update. Electrophoresis 19:1960-1970.

Torella,F.,andR.Y.Morita.1981. Microcultural study of bacterial size changes and microcolony and ultramicrocolony formation by heterotrophic bacteria in seawater. Applied and Environmental Microbiology 41:518-527.

Tweeddale, H., L. Notley-McRobb, and T. Ferenci. 1998. Effect of slow growth on metabolism of Escherichia coli, as revealed by global metabolite pool ("metabolome") analysis. Journal of Bacteriology. 180:5109-5116. VanBogelen, R. A., E. R. Olson, B. L. Wanner and F. C. Neidhardt 1996. Global analysis of proteins synthesized during phosphorus restriction in Escherichia coli. Journal of Bacteriology 178:4344-66.

Vancanneyt, M., F. Schut, C. Snauwaert, J. Goris, J. Swings, and J. C. Gottschal. 2001. Sphingomonas alaskensis sp. nov., a dominant bacterium from a marine oligotrophic environment. International Journal of Systematic and Evolutionary Microbiology 51:73-79.

Vaulot, D., D. Marie, R. J. Olson, and S. W. Chisholm. 1995. Gowth of Prochlorococcus. A photosynthetic prokaryote, in the equatorial Pacific Ocean. Science 268:1480-1482.

223 Velicer, G. J., T. M. Schmidt, and R. E. Lenski. 1999. Application of traditional and phylogenetically based comparative methods to test for a trade-off in bacterial growth rate at low versus high substrate concentration. Microbial Ecology 38:191-200.

Velicer, G. J., T. M. Schmidt and R. E. Lenski 1999. Application of traditional and phylogenetically based comparative methods to test for a trade-off in bacterial growth rate at low versus high substrate concentration. Microbial Ecology 38: 191-200.

Velimirov, B. 2001. Nanobacteria, ultramicrobacteria and starvation forms: a search for the smallest metabolizing bacterium. Microbes and Environments 16: 67- 77.

Volkert M. R., P. C. Loewen, J. Switala, D. Crowley and M. Conley. 1994. The (argF-lacZ)205(U169) deletion greatly enhances resistance to hydrogen peroxide in stationary-phase Escherichia coli. Journal of Bacteriology. 176:1297-1302.

Vybiral,D.,E.B.M.Denner,C.M.Haller,H.J.Busse,A.Witte,M.G.Hofle, and B. Velimirov. 1999. Polyphasic classification of 0.2 mu m filterable bacteria from the western Mediterranean Sea. Systematic and Applied Microbiology 22:635-646.

Wang, Y., P. C. K. Lau, and D. K. Button. 1996. A marine oligobacterium harboring genes known to be part of aromatic hydrocarbon degradation pathways of soil Pseudomonads. Applied and Environmental Microbiology 62:2169-2173.

Wase D.A. and Y.R. Patel. 1985. Variations in the volumes of microbial cells with change in the agitation rates of chemostat cultures. Journal of General Microbiology 131: 725–736.

Wasinger, V. C., B. Bjellqvist, and I. Humprey-Smith. 1997. Proteomic 'contigs' of Ochrobactrum anthropi, application of extensive pH gradient. Electrophoresis 18:1373-1383.

Wasinger,V.C.,S.J.Cordwell,A.Cerpa-Poljak,J.X.Yan,A.A.Gooley,M.R. Wilkins, M. W. Duncan, R. Harris, K. W. Williams, and I. Humphrey- Smith. 1995. Progress with the gene product mapping of the Mollicutes Mycoplasma genitalium. Electrophoresis 16:1090-1094.

Wasinger, V. C., D. J. Pollack, and I. Humprey-Smith. 2000. The proteome of Mycoplasma genitalium. European Journal of Biochemistry 267:1571-1582.

Wayne, L. G., D. J. Brenner, R. R. Colwell, P. A. D. Grimont, O. Kandler, M. I. Krichevsky, L. H. Moore, W. E. C. Moore, R. G. E. Murray, E. Stackebrandt, M. P. Star, and H. G. Truper. 1987. Report of the Ad Hoc Committee on Reconciliation of Approaches to Bacterial Systematics. International Journal of Systematic Bacteriology 37:463-464.

224 West, N.J., and Scanlan, D.J. 1999 Niche-partitioning of Prochlorococcus populations in a stratified water column in the eastern North Atlantic Ocean. Applied and Environmental Microbiology 65:2585-2591.

West, N.J., Schonhuber, W.A., Fuller, N.J., Amann, R.I., Rippka, R., Post, A.F., and Scanlan, D.J. 2001 Closely related Prochlorococcus genotypes show remarkably different depth distributions in two oceanic regions as revealed by in situ hybridization using 16S rRNA-targeted oligonucleotides. Microbiology 147: 1731-1744.

White, A. L., and L. S. Jahnke. 2002. Contrasting effects of UV-A and UV-B on photosynthesis and photoprotection of -carotene in two Dunaliella spp. Plant Cell Physiology. 43:877-884.

Whitman, W. B., D. C. Coleman, and W. J. Wiebe. 1998. Prokaryotes: The unseen majority. Proceedings of the National and Academy of Science USA 95:6578-6583.

Wiebinga, C. J., M. J. W. Veldhuis, and H. J. W. de Baar. 1997. Abundance and productivity of bacterioplankton in relation to seasonal upwelling in the nothwest Indian Ocean. Deep-Sea Research I 44:451-476.

Wignall, P. B. and R. J. Twitchett. 1996. Oceanic anoxia and the end permian mass extinction. Science 272:1155-1158. Wilhelm,E.R.,R.BattinoandR.J.Wilcock.1977. Low pressure solubility of gases in liquid water. Chemical Reviews. 77:219-262.

Wilkins, M. R., E. Gasteiger, C. H. Wheeler, I. Lindskog, J. C. Sanchez, A. Bairoch, R. D. Appel, M. J. Dunn and D. F. Hochstrasser. 1998. Multiple parameter cross-species protein identification using MultiIdent - a World-Wide Web accessible tool. Electrophoresis 19:3199-3206.

Yabuuchi, E., I. Yano, H. Oyaizu, Y. Hashimoto, T. Ezaki, and H. Yamamoto. 1990. Proposal of Sphingomonas paucimobilis gen nov and comb nov, Sphingomonas parapaucimobilis sp nov, Sphingomonas yanoikuyae sp nov, Sphingomonas adhesiva sp nov, Sphingomonas capsulata comb nov, and two genospecies of the genus Sphingomonas. Microbiology and Immunology 34:99-119.

Yamazaki, M., L. Thorne, M. Mikolajczak, R. W. Armentrout, and T. J. Pollock. 1996. Linkage of genes essential for synthesis of a polysaccharide capsule in Sphingomonas strain S88. Journal of Bacteriology 178: 2676–2687.

You, Y., N. Suginohara, M. Fukasawa, I. Yasuda, I. Kaneko, H. Yoritaka, and M. Kawamiya. 2000. Role of the Okhotsk Sea and Gulf of Alaska in forming the North Pacific Intermediate Water. Journal of Geophysical Research 105:3253-3280.

225 Zgurskaya, H. I., M. Keyhan, and A. Matin. 1997. The sigma S level in starving Escherichia coli cells increases solely as a result of its increased stability, despite decreased synthesis. Molecular Microbiology. 24:643-651.

Zubkov, M. V., B. M. Fuchs, G. A. Tarran, P. H. Burkhill and R. Amann. 2003. High rate of uptake of organic nitrogen compounds by Prochlorococcus cyanobacteria as a key to their dominance in oligotrophic oceanic waters. Applied and Environmental Microbiology 69:1299-1304.

Zubkov, M. V., B. M. Fuchs, P. H. Burkill, and R. Amann. 2001. Comparison of cellular and biomass specific activities of dominant bacterioplankton groups in stratified waters of the Celtic Sea. Applied and Environmental Microbiology

226